Home

Graphical User Interface User Manual

Contents

1. io columnSubset_dataset csv columnSubset_optAndStruct csv b opt_and_struct csv csv XOr cSV csv Last Access 2010 11 30 2010 11 30 2010 11 30 2010 11 30 As clearly visible in Fig 17 the Configuration panel shows the list of columns originally present in the input data file that can be selected by proper check boxes Note that the whole content of the data file in principle a massive data set is not shown but simply labelled by column meta data as originally present in the file X Dele X X X x Fig 18 The Feature Selection operation the new file created 3 5 2 2 Column Ordering RESOURCE MANAGER File Editor EJ Workspace myfirstWS File opt_and_struct_csv lt Operation Feature Selection Columns Ordering Sort Rows by Column Column Shuffle Row Shuffle Split by Rows Dataset Scale Single Column Scale Description It creates a new file containing only the selected columns It creates a new file containing the new ordering of columns as specified in the column fields It creates a new file containing the sorted rows by specifying the column reference index It creates a new file containing shuffled columns It creates a new file containing shuffled rows It creates as many new files as desired each containing the specified percentage of rows The rows distribution can be randomly extracted Use only numerical Dataset It creates a new file contain
2. columns Range 1 A It creates a new file containing the new ordering of 2 col Float _ 0 rosi mar columns as specified in the column fields 1 mein 5 3 col3 Float It creates a new file containing the sorted rows by Output ey eee oy ee specifying the column reference index 4 col Float File optAndStruct Column Shuffle It creates a new file containing shuffled columns 5 cols Float Name Row Shuffle It creates a new file containing shuffled rows 6 cob Float It creates as many new files as desired each 7 col7 Float Split by Rows containing the specified percentage of rows The 8 cols Float rows distribution can be randomly extracted Use only numerical Dataset 9 coe Double Dataset Scale t creates a new file containing all data scaled in the 10 colo Double selected range 1 1 or 0 1 11 col11 Double It creates a new file containing a single column data scaled in the selected range 1 1 or 0 1 12 coli2 Short Single Column Scale leaving unmodified the other columns in the original order Program DAta Mining amp Exploration Fig 31 The Dataset Scale operation step 1 3 5 2 8 Single Column Scale This dataset operation that works on numerical data files only permits to normalize a single selected column between those contained in the original file in one of two possible ranges respectively 1 1 or 0 1 The result is the creation of a new file of the same type and with the
3. number of iterations error tolerance training mode 1 MSE Batch 2 MSE Incremental Fig 44 The configuration options in the Train use case Test a sort of validation of the training phase It can done by submitting the same training dataset or a subset or a mix between already submitted and new dataset patterns RESOURCE MANAGER File Editor EJ Experiment Setup Select a Regression MLP m Select a Running T Functionality Mode pa Field is Required Test Set v Network File Fig 45 The configuration options in the Test use case Run normal use of the already trained model RESOURCE MANAGER File Editor Experiment Setup Select a Select a Running R MLP R Functionality so x Mode 7 x Field is Required Run Set v Network File Submit Fig 46 The configuration options in the Run use case Full the complete and automatic serialized execution of the three previous use cases train test and Run It is a sort of workflow considered as a complete and exhaustive experiment for a specific problem RESOURCE MANAGER File Editor 3 Experiment Setup Select a Regression _MLP v Select a Running datati Functionality Mode Field is Required Train Set v Validation Set Test Set v Network File v number of input nodes number of nodes for hidden layer number of output nodes number of iterations error
4. YLabel Y xAxis Z E Flip Marker_Shape X v _ yAxis u g v E Flip Linear Correlation Clear tab mA Piot H Save Plot as cgil Export in another window Fig 35 The Scatter 2D tab As shown in Fig 35 there is the possibility to create and visualize a scatter 2D of any table file previously loaded or produced in the web application There are several options e Workspace the user workspace hosting the table Table the name of the table to be plotted xAxis selection of the x column of table to be plotted yAxis selection of the y column of table to be plotted Marker_Size size of the marker Color color of the plot bars Marker_Shape shape of the marker Line_Width width of the bars Linear Correlation enable the drawing of a line based on linear correlation of columns Flip enable the flipping of the x Axis of the plot Flip enable the flipping of the y Axis of the plot Title title of the plot Xlabel label of the x axis Ylabel label of the y axis Grid enable disable the grid in the plot Clear Tab clear the tab Plot creation and visualization of the selected histogram Save Plot As plot saving with user typed name Export in another window the plot will be moved in an independent tab of the web browser Add Tab enable the creation of a multi layer scatter 2D as shown in Fig 36 21 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights
5. New Workspace at left corner of workspace manager window 12 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program As consequence the user must assign a name to the new workspace by filling in the form field as in Fig 9 Hew Workspace Name OK Cancel Fig 9 the form field that appears after pressing the New Workspace button After creation the active workspace can be populated by data and experiments Fig 10 DAME Application User bresciamax gmail com LogOut amp Model Manuals Other Services v Science Cases z RESOURCE MANAGER Workspace V File Manager Workspace E New Workspace Ulo Plot Editor ta Image Viewer imipgnaExp a x T E Dow Edit File Type Last Access M Dele Rename Workspace F Upload Experiment Delete di A B amp dataset_run_20 ascii ascii 2013 06 27 x TF P primo C x R o amp dataset test 20 ascii ascii 2011 12 06 x f mipqnaExp B w x R B dataset train_80 ascii ascii 2011 12 06 x 7 mipexp mipana xe a x L 3 a dataset train_80 ascii mipgna_TRAIN_frr ASCII 2013 07 02 x 4 phat z ad x B mipgna_TRAIN_weights txt ASCII 2013 07 02 x f clashRegr B dd x amp amp opt_and struct csv csv 2013 07 02 se tf 4 tx B E x g B xor csv csv 2011 12 05 x GG 4 img1 B d x J V My Experiments
6. Program File Manager Area the portion of Resource Manager tab dedicated to list the data files belonging to various workspaces All files present in this area are considered as input files for any kind of experiment Download command When selected the user can download locally on his HD the selected file Dataset Editor command When selected a new tab is open where the user can create editdataset files by using all available dataset manipulation features Delete file command When selected the user can delete the selected file from current workspace Experiment List Area The portion of Resource Manager tab dedicated to the list of experiments and related output files present in the selected workspace Experiment verbose list command When selected the user can open the experiment file list for experiment in ended state in a verbose mode showing all related files created and stored Delete Experiment command by clicking on it the entire experiment all listed files is erased Download experiment file command When selected the user can download locally on his HD the related experiment output file AddinWS command When selected the related file is automatically copied from the Experiment List Area to the currently active workspace File Manager Area This feature is useful to re use an output file of a previous experiment as input file of a new experiment in the figure look at the file weights txt that after this command is also lis
7. bestarsqna x Workspace jmipqnaExp P 4survey B a x Experiment Status Last Access SM Delete PI gt gt classtrain1 ended 2013 07 02 a f FMLPGAExp B cd x i gt classtest1 ended 2013 07 02 x b classrunt ended 2013 07 02 x b classfulli ended 2013 07 02 x b regrtrain1 ended 2013 07 02 x b regrtest1 ended 2013 07 02 x f b regrrunt ended 2013 07 02 x E Fig 10 the active workspace created in the Workspace List Area 3 4 Header Area At the top segment of the DMS GUI there is the so called Header Area Apart from the DAME logo it includes a persistent menu of options directly related to information and documentation this document also available online and or addressable through specific DAME program website pages 13 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program DAME Application User bresciarnax gmail com Logout Me A i mm App Manuals 2 ip ya LI Reterence Guide RI ESOM Manual PhotoRApToR LI Photometric Recshifts POE amp Lectures va KR GUIUser Manual Re FMLPGA Manual D STraDivt El Photometric Quasars E Science Production E YouTube Channel j ALS d KR Add Your Model O GAME Manual LI VOGCIusters O Giobular Ciusters LI Nevesiatters Ee Official Website Ee K Means Manual D KNME with DAME O AGN Classification LI Release Notes ER Citation Policy RESOURCE MANA
8. 03 06 2009 22 03 2010 07 07 2007 02 10 2007 20 02 2008 05 01 2009 30 11 2008 14 04 2010 28 10 2010 30 11 2010 30 11 2010 30 11 2010 42 DAta Mining amp Exploration Program Acknowledgments project VOTECH Virtual Observatory Technological Infrastructures and by the Italian PON ik DAME program has been funded by the Italian Ministry of Foreign Affairs the European S Co P E Leaders of the project are prof G Longo and prof G S Djorgovski The current release of the data mining Suite is a miracle due mainly to the incredible effort of in alphabetical order Giovanni Albano Stefano Cavuoti Giovanni d Angelo Alessandro Di Guido Francesco Esposito Pamela Esposito Michelangelo Fiore Mauro Garofalo Marisa Guglielmo Omar Laurino Francesco Manna Alfonso Nocella Sandro Riccardi Bojan Skordovski Civita Vellucci We want to really thank all actors who contribute and sustain our common efforts to make the whole DAME Program a reality coming from University Federico II of Naples INAF Astronomical Observatory of Capodimonte and Californian Institute of Technology Max 000 43 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program DAME Program we make science discovery happen oe ANT Pi Ji Sr 44 DAMEWARE GUI User Manual This document contains proprietary inf
9. A STE Data files ue upload download files from to remote URI gt 2 upload edit Dataset files edit data files data files editor create and launch an experiment navigate through experiments move files from experiment to workspace 3 create configure experiments List of input files configuration files output files List of input files configuration files output files experiment experiment 4 launch experiments N plots RS visualization Dr statistics N N 6 Explore results 5 store results 5 store results Fig 7 The right sequence to configure and execute an experiment workflow So far the basic role of a workspace is to make easier to the user the organization of experiments and related input output files For example the user could envelope in a same workspace all experiments related to a particular functionality domain although using different models It is always possible to move copy files from experiment to workspace list in order to re use a same dataset file for multiple experiment sessions i e to perform a workflow After access the user must select the active workspace If no workspaces are present the user must create a new one otherwise the user must select one of the listed workspace The user can always create a new workspace by pressing the button as in Fig 8 Workspace Manager z Mew Workspace o Workspace Fig 8 the button
10. Fig 4 confirming the correct coming up of the activation procedure Registering on DAME Posta in arrivo x X FEregistration dame it a me Dear max cipolla Thank you for registering to DAME application suite Your account will be activated as soon as possible Your username is bresciamax gmail com Your password is Please visit http voneural na infn it for all information about DAME services For any request please contact by e mail helpdame gmail com by Skype helpdame The DAME Staff DAME Data Mining amp Exploration Program Fig 4 An example of e mail received by the user after submission of registration info After that the user must wait for a second e mail which will be the final confirmation about the activation of the account This is required in order to provide an higher security level Once the user has received and password The webapp will appear as the activation confirmation he can access the webapp by inserting e mail address shown in Fig 5 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program DAME Application User bresciamax gmail com LogOut E App Manuals v Model Manuals v Other Services v Science Cases v RESOURCE MANAGER Workspace v File Manager Workspace Zz New Workspace M Plot Editor m image Viewer FMLPGAExp E Dow Edit File Type La
11. MLP weight matrix file output of the training phase to be re used as input weight matrix of a test or validation session of the same network In order to make available this fundamental feature in our application the icon command nr 18 AddInWS in Fig 6 associated to each output file of an experiment can be selected by the user in order to copy the file from experiment output list to the workspace input list becoming immediately available as input file for new experiments belonging to the same workspace _as important remark in the beta release it is not yet possible to move files from a workspace to another The alternative procedure to perform this action is to download the file on user local Hard Disk and to upload it into another desired workspace in the webapp 3 6 Plotting and Visualization The final release of the web application offers two new options for plotting and visualization of data tables or images 3 6 1 Plotting By pressing the Plot Editor button in the main menu a series of plot tabs will appear Each one is dedicated to a specific type of plot for instance Histogram Scatter Plot 2D Scatter Plot 3D and Line Plot DAME Application User jobfmann gmail com LogOut L me App Manuals v Model Manuals v Cloud Services v Science Cases v Graphic View Histogram 400 350 300 gt 250 200 150 100 50 0 ga 02 01 00 01I 02 03 04 05 068 07 0U 09 r0 at ae 13 149 195 1868 17 1
12. User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program 3 7 1 Re use of already trained networks In the previous section a general description of experiment use cases has been reported A specific more detailed information is required by the Run use case As known this is the use case selected when a network for example the MLP model has been already trained i e after training use case already executed RESOURCE MANAGER Experiment Setup DI Selectaln oo Select a Running __ Classification MLP Train Functionality Ta Mi n x Field is Required Train Set xor csy r Validation Set Ww Network File n number of input nodes 2 number of nodes for hidden layer 2 number of output nodes number of iterations error tolerance training mode Submit Fig 50 An example of Classification MLP training case for the XOR problem Note mM i j Experiment Finished Please refer to MyFirst WS workspace for results Fig 51 The popup status at the end of the XOR problem experiment 36 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program o My Experiments Workspace hiyFirstWS Experiment Status Last Access Delete al xorTrain ended 2013 09 09 x C Download Addinws File Type Desc
13. in by the user all fields are required e Name of the user e Family name of the user e User e mail the user e mail it will become his access login It is important to define a real address because it will be also used by the DMS for communications feedbacks and activation instructions e Country country of the user e Affiliation the institute academy society of the user e Password a safe password at least 6 chars without spaces and special chars DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program DAME Registration Window DAME Registration Form Name Family Name Email Confirm Email Country Affiliation Password Confirm Password novative User Mail i I Eyen a minin address ta Sets New on DAME WebApp ipinin Revistertow Register Now Allowed range 6 16 You can obtain the access by following characters a simple registration procedure 1 Compile the registration form click Register Now button 2 immediately after you will receive by e mail a welcome message 3 Check for an e mail message with your account confirmation 4 Go back at this page and sign in i Vr 3 imetamnt INAF OACN di Fig 3 The user registration form After submission an e mail will be immediately sent at the defined address
14. learning rule Associated functionalities are classification and regression This model is available in two versions CPU and GPU the second one 1s the parallelized version e SVM Support Vector Machine model Associated functionalities are classification and regression e MLPQNA Multilayer Perceptron neural network trained by Quasi Newton learning rule Associated functionalities are classification and regression e LEMON Multilayer Perceptron neural network trained by Levenberg Marquardt learning rule Associated functionalities are classification and regression e RANDOM FOREST Randomly generated forest of decision trees network Associated functionalities are classification and regression KMEANS Standard Kmeans algorithm Associated functionality is clustering CSOM Customized Self Organizing Feature Map for clustering on FITS images GSOM Gated Self Organizing Map SOFM for clustering on text and or image files PPS Probabilistic Principal Surfaces for feature extraction SOM Self organizing Map for pre clustering on text or image files SOM Auto SOM with an automatized post processing phase for clustering on text or image files SOM Kmeans SOM with a Kmeans based post processing phase for clustering on text or image files e SOM TWL SOM with a Two Winners Linkage TWL based post processing phase for clustering on text or image files e SOM UmatCC SOM with an U matrix Connected Components UmatCC based post pr
15. other columns in the original order Fig 29 The Split by Rows operation step 1 v File Manager Workspace myfirstWS C Dow amp Edit File Type Last Access X Dele ee amp opt_and_struct csv csv 2010 11 30 x os rowShuffle_optAndStruct csv 2010 12 01 x i l amp rowSort_optAndStruct csv 2010 12 01 x o shuffle_optAndStruct csv 2010 12 01 x E l amp spiit30_optAndStruct csv 2010 12 01 x amp split70_optAndStruct csv 2010 12 01 x lt a succ split70_optAndStruct x Fig 30 The Split by Rows operation the new files created 3 5 2 7 Dataset Scale This dataset operation that works on numerical data files only permits to normalize column data in one of two possible ranges respectively 1 1 or 0 1 This is particularly frequent in machine learning experiments to submit normalized data in order to achieve a correct training of internal patterns The result is the creation of a new file of the same type and with the same extension of the original file named as scale_ lt user selected name gt 1 e with specific prefixscale Details are reported in Fig 31 23 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved RESOURCE MANAGER File Editor Workspace myfirstWS Configuration File opt_and_struct_csv 9 lt Operation Description 4 OK Casio oe t creates a new file containing only the selected 1 ciali Float
16. same extension of the original file named as scaleOneCol lt user selected name gt i e with specific prefixscaleOneCol Details of the simple procedure are reported in Fig 32 RESOURCE MANAGER File Editor 3 Workspace myfirstWS Configuration File opt_and_struct_csv 9 lt Operation Description 4 OK Feature Selection It creates a new file containing only the selected 1 col Float OR columns Range 1 It creates a new file containing the new ordering of 2 col2 Float _ 0 a ae columns as specified in the column fields 1 cedola see 3 col Float i t creates a new file containing the sorted rows by Select x Sort Rows by Coluna specifying the column reference index 4 col4 Float index Column Shuffle It creates a new file containing shuffled columns 5 cols Float Output sa File optAndStruct Row Shuffle t creates a new file containing shuffled rows 6 cob Float Name It creates as many new files as desired each 7 col Float Split by Rows containing the specified percentage of rows The 8 cols Float rows distribution can be randomly extracted Use only numerical Dataset 9 cols Double Dataset Scale It creates a new file containing all data scaled in the 10 col10 Double selected range 1 1 or 0 1 ae a 11 col11 Double It creates a new file containing a single column data scaled in the selected range 1 1 or 0 1 12 col12 Short na oa See leaving unmodified the other columns in the original
17. sarcasmo esl aol Eunocmeont Cas Gauratinn Bile Fig 6 The Web Application main areas and commands 11 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program 3 3 Workspace Management A workspace is namely a working session in which the user can enclose resources related to scientific data mining experiments Resources can be data files uploaded in the workspace by the user files resulting from some manipulations of these data files i e dataset files containing subsets of data files selected by the user as input files for his experiments eventually normalized or re organized in some way see section3 5 for details Resources can also be output files i e obtained as results of one or more experiments configured and executed in the current active workspace see section 3 7for details The user can create a new or select an existing workspace by specifying its name After opening the workspace this automatically becomes the active workspace This means that any further action manipulating files configuring and executing experiments upload download files will result in the active workspace Fig 7 In this figure it is also shown the right sequence of main actions in order to operate an experiment workflow in the correct way Workspace 1 create open a List of k Experiments The user can wy
18. through the GUI options and features In other words to show how to navigate and to interact with the application interface in order to create working spaces experiments to upload download and edit data files We will stop our discussion here at level of configuration of the models for which specific manuals are available This in order to separate the use of the GUI from the theoretical implications related to the setup and use of available data mining models The access gateway its complete documentation package and other resources is at the following address http dame dsf unina it dameware html DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program Last pages of this document host tables with Abbreviations amp Acronyms Reference and Applicable document lists and the acknowledgments All over the document the references are labeled as Rxx for Reference documents and Axx for Applicable documents xx is the incremental index as reported in the list tables Applicable documents are not public references technical documents internal to the DAME working group included for quick technical references Users external to the working group may ask to consult privately these documents by e mail motivating the reasons The complete list of the internal documentation is available at the following addre
19. to a workspace previously created by the user All these data are listed in the Files Manager sub window These data can be in one of the supported formats i e data formats recognized by the web application as correct types that can be submitted to machine learning models to perform experiments They are FITS tabular and image fits files ASCII txt or dat ordinary files VOTable VO compliant XML document files CSV Comma Separated Values csv files JPEG GIF and PNG images The user has to pay attention to use input data in one of these supported formats in order to launch experiments in a right way Other data types are permitted but not as input to experiments For example log jpeg or not supported text files are generated as output of experiments but only supported types can be eventually re used as input data for experiments There is an exception to this rule for file format with extension ARFF Attribute Relation File Format These files can be uploaded and also edited by dataset editor by using the type CSV But their extension ARFF is considered unsupported by the system so you can use any of the dataset editor options to change the extension automatically assigned as CSV Then you can use such files as input for experiments These output file are generally listed in the Experiment Manager sub window that can be verbosely open by the user by selecting any experiment when it is und
20. tolerance training mode 1 MSE Batch 2 MSE Incremental 34 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved E f xploration Wr hd Cd Fig 47 The configuration options in the Full use case In all the above use case tabs the help button redirects to a specific web page reporting in verbose mode detailed description of all parameters In particular the parameter fields marked by an asterisk are considered required by the user All other parameters can be left empty by assuming a default value also reported in the hep page Functionality Regression with MLP Parameter specifications Use Case TRAIN Fig 48 Example of a web page automatically open after the click on the help button After completion of the parameter configuration the Submit button launches the experiment After launch of an experiment it can result in one of the following states Enqueued the execution is put in the job queue Running the experiment has been launched and it is running Failed the experiment has been stopped or concluded with any error occurred Ended the experiment has been successfully concluded w Experiment Manager Experiment Status Last Access MX Vette gt myfirstExp ended 2010 06 23 x gt myClass xp running 2010 06 24 x Fig 49 Some different state of two concurrent experiments 35 DAMEWARE GUI
21. 1 Rel2 0 statement_of_work_VONEURAL SOW NA 0001 Rel1 0 MLP_user_manual_VONEURAL MAN NA 0001 Rel1 0 pipeline_test_VONEURAL PRO NA 0001 Rel 1 0 scientific_example VONEURAL PRO NA 0002 Rel 1 1 frontend_VONEURAL SDD NA 0004 Rel1 4 FW_VONEURAL SDD NA 0005 Rel2 0 REDB_VONEURAL SDD NA 0006 Rel1 5 driver VONEURAL SDD NA 0007 Rel0 6 dm model_ VONEURAL SDD NA 0008 Rel2 0 ConfusionMatrixLib_VONEURAL SPE NA 0001 Rel1 0 softmax_entropy_ VONEURAL SPE NA 0004 Rel1 0 VONeuralMLP2 0_VONEURAL SPE NA 0007 Rel1 0 dm_model VONEURAL SRS NA 0005 Rel0 4 FANN_MLP_VONEURAL TRE NA 0011 Rel1 0 DMPlugins_DAME TRE NA 0016 Rel0 3 BetaRelease_ReferenceGuide DAME MAN NA 0009 Rel1 0 BetaRelease_ Model MLP_UserManual DAME MAN NA 0011 Rell 0 BetaRelease_ Model SVM_UserManual DAME MAN NA 0013 Rel1 0 BetaRelease_ Model MLPGA_UserManual DAME MAN NA 0012 Rel1 0 Author DAME Working Group Brescia Brescia DAME Working Group D Abrusco D Abrusco Cavuoti Manna Fiore Nocella d Angelo Cavuoti Di Guido Cavuoti Skordovski Skordovski Cavuoti Skordovski Laurino Di Guido Brescia Cavuoti Brescia Cavuoti Brescia Cavuoti Brescia Brescia Tab 4 Applicable Documents DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program Date 15 10 2008 19 02 2008 30 05 2007 12 10 2007 17 07 2007 06 10 2007 18 03 2009 14 04 2010 29 03 2010
22. 89 19 20 ZI 22 23 e x V Graphic Options Add Tab Bin Placement Width 0 04 Title Histogram Main XLabel X Workspace WS20111221 v Bar_Style ily Filled v YLabel Y Table dataset_test_20 ascii v Color E s 7 Grid xAxis g r v E Flip Line_Width 1 v Clear tab m Plot H Save Plot as pl Export in another window Fig 33 The Histogram tab 25 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program As shown in Fig 33 there is the possibility to create and visualize an histogram of any table file previously loaded or produced in the web application There are several options e Workspace the user workspace hosting the table Table the name of the table to be plotted xAxis selection of the column of table to be plotted Bar_Style style of the bars Color color of the plot bars Line_Width width of the bars Flip enable the flipping of the x Axis of the plot Title title of the plot Xlabel label of the x axis Ylabel label of the y axis Grid enable disable the grid in the plot Bin Placement change the bin width of plot Clear Tab clear the tab Plot creation and visualization of the selected histogram Save Plot As plot saving with user typed name Export in another window the plot will be moved in an independent tab of the web browser Add Tab enable the creation of a multi layer histogram a
23. A IEEE INAF JPEG LAR MDS MLP MLPGA MLPQNA MSE NN OAC PC PI REDB RIA SDSS SL SOFM SOM SW UI URI VO XML DAta Mining amp Exploration Program Meaning Institute of Electrical and Electronic Engineers IstitutoNazionale di Astrofisica Joint Photographic Experts Group Layered Application Architecture Massive Data Sets Multi Layer Perceptron MLP with Genetic Algorithms MLP with Quasi Newton Mean Square Error Neural Network Osservatorio Astronomico di Capodimonte Personal Computer Principal Investigator Registry amp Database Rich Internet Application Sloan Digital Sky Survey Service Layer Self Organizing Feature Maps Self Organizing Maps Software User Interface Uniform Resource Indicator Virtual Observatory eXtensible Markup Language Tab 2 Abbreviations and acronyms 40 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved Reference amp Applicable Documents ID RI R2 R3 R4 RS R6 R7 R8 R9 R10 R11 R12 Title Code The Use of Multiple Measurements in Taxonomic Problems in Annals of Eugenics 7 p 179 188 Neural Networks Oxford University Press GB for Pattern Recognition Neural Computation Data Mining Introductory and Advanced Topics Prentice Hall Mining the SDSS archive I Photometric Redshifts in the Nearby Universe Astrophysical Journal Vo
24. DAta Mining amp Exploration Pa LI ET Dipartimento di Scienze Fisiche 2 istmuronazionate aiastrorisica 9 CALTECH Nore 7 gt a gt PRI S Universita di Napoli Federico H Tadeo OSSERVATORIO ASTRONOMICO di CAPODIMONTE Q Ny Graphical User Interface User Manual DAME MAN NA 0010 Issue 1 4 Date March 16 2015 Author M Brescia S Cavuoti Doc GUL UserManual DAME MAN NA 0010 Rel1 4 IMAF DACH DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program DAME Program we make science discovery happen INDEX De RAAT OIC 1h ae E oa 5 2 PWT e E E ces icuaietousseensaiaetetencecbremesesent 5 EU EIA 6 LL MOSER 1S Gi alli A ACCESS inni 7 S MANE OTA TC ONS RR RR IE 10 zr Workspace Mana eno M essorer oiee ear EO Eae arrera 12 acc ia na a A lira paia S E O 13 F Da Mansioni 15 JDL Upload erdal alianti lecita 16 3 5 2 How to Create dataset files seahecacesiusousacvanandasnnsacedatecacssbubles destanesdnvasaacedatedecesbubdedecvanaedauasiacedatenes 17 3 5 2 1 Fe AUS ie EC 1 FON sessioissa i 18 SZ CL RR REA 19 3 5 2 3 Sor Rows Dy C O MITI seisessssnieeper teste eres chet ete eset mec sated ERO A ENTEN EN REEE 20 3 5 2 4 CRI RR E 21 3 5 2 5 Rolli era 22 3 5 2 6 PECOS IROWS aa E E EE ENE N EE 23 ENN MAE ASC 6 E e E TI 23 3 5 2 8 SALO CAS PENE RE 24 Bide Downoad Tsen eE ea E E E R 24 354 Mon
25. DAta Mining amp Exploration Program 1 Introduction package The final release arises from the very primer version of the web application a release which has been made available to public domain since July 2010 Currently this release has been more updated by fixing residual bugs and by adding more functionalities and models The final developing team has spent much efforts to fix bugs satisfy testing user requirements suggestions and to improve the application features by integrating several other data mining models always coming from machine learning theory which have been scientifically validated by applying them offline in several practical astrophysical cases photometric redshifts quasar candidate selection globular cluster search transient discovery etc All cases dealing with time domain data rich astronomy In this scenario the a release has covered the role of an advanced prototype useful to evaluate tune and improve main features of the web application basically in terms of ik present document is part of the DAMEWARE Web Application Suite user side documentation e User friendliness by taking care of the impact on new users not necessarily expert in data mining or skilled in machine learning methodologies by paying particular attention to the easiness of navigation through GUI options and to the learning speed in terms of experiment selection preprocessing setup and execution e Data I O handling easiness to uplo
26. GER Gi MLPBP Manual LA Sky Transients Write Us Workspace DI MLPONA Manual v File Manager E About Us E PPS Manual Workspace B New V orkspace lis Piot Editor lii r miponaExp Ee SOFM Manuals j di E Dow Edt Fie Type LastAccess M Dele p Rename Workspace Ee SOM Manual it 3 Delete KA SVM Manual a dataset_run_20 0sc asc 2013 08 27 x Fig 11 The GUI Header Area with all submenus open The options are described in the following table Tab 1 OPTIONS HEADER Application Manuals DESCRIPTION http dame dsf unina it dameware html appman http dame dsf unina it dameware html plugin E http dame dsf unina it dameware htmlmanuals http dame dsf unina it dameware htmlmanuals T http dame dsf unina it dameware html manuals http dame dsf unina it dameware html manuals anua Model Manuals http dame dsf unina it dameware htmlmanuals http dame dsf unina it dameware htmlfmanuals PPS Manual http dame dsf unina it dameware htmlmanuals SOFM Manual bttp dame dsf unina it dameware html manuals SOM Manual http dame dsf unina it dameware htmlfmanuals SVM Manual http dame dsf unina it dameware html manuals PhotoRApToR http dame dsf unina it dame_photoz html photoraptor __ nup gdame dst unina Iy dame xappa numI STraDIWA http dame dsf unina it dame_td html VOGCLUSTERS App OMEN SERVICES http dame dsf unina it vogclusters html KNIME with DAME http dame dsf unin
27. Reserved DAta Mining amp Exploration Program Model Manuals Cloud Services RESOURCE MANAGER Histogram E Scatter Plot 2D E Scatter Plot 3D E Line Plot EJ Graphic View Multi_Scatter_Plot_2D WOTEB Mn A GB B8 BB c l Title Multi Scatter Plot 2D Workspace WS20111221 v Marker_Size 2 m XLabel X Table dataset_test_20 ascii y Color N YLabel Y Fig 36 A multi layer scatter 2D plot ADVERTISEMENT whenever the user change any parameter of the current plot it is needed to click the button Plot to refresh the visualized plot DAME Application User jobfmann gmail com LogOut L Model Manuals Cloud Services RESOURCE MANAGER Histogram E Scatter Plot 2D 9 Scatter Plot 3D E Line Plot Graphic View Scatter Plot 3D si Add Tab Main a 2 z 2 Workspace Manual v Marker_Size 1 XLabel X Theta 15 Table train csv v Color II YLabel Y F Grid oie Wa E Fip Marker_Shape A x ZLabel Z Foo yAxis Z v E Flip zAxis u g v E Ap Clear tab m Piot Save Plot as GG sportin another window Fig 37 The Scatter Plot 3D tab As shown in Fig 37 there is the possibil
28. UI User Manual This document contains proprietary information of DAME project Board All Rights Reserved v File Manager Workspace mipExp Dow Edit File Ei mip_TRAIN_weights D xor csv et oo xor_run csv v My Experiments Workspace mipExp Experiment 4 xorTrain Download Addnws eo Cy Ual Status ended File mip_TRAIN_errorPiot jpeg mip_TRAIN_error csv mip_TRAIN log mip_TRAIN_tmp_weights mip_TRAIN_weights Type mip CSV Mining amp Exploration Program Last Access x Dele 2011 05 30 x 2011 05 30 x 2011 05 30 x Last Access XM Delete 2011 05 30 x Type Description peg scatter plot of the epochs vs error csv epoch error file txt logfile mip _nettmp file mip trained network file Fig 54 The operation to move the trained network file in the Workspace input file list To do this the output weight file must become an input file in the workspace file list as already explained in section 3 5 4 otherwise it cannot be used as input of Test Run use case experiment Fig 54 Also the workspace currently active hosting the experiment we are going to do must contain a proper input file for Run cases 1 e without target columns inside So far the second step is to populate the workspace file list with trained network and Test Run compliant input files and then to configure and execute the test experiment see Fig 55 RESOURCE MANAGER
29. Upload in mipExp Workspace EJ Workspace mipExp Experiment xorTest Select a Classification_MLP v Functionality Select a Runnin 9 Test Mode Field is Required Experiment Setup Test Set xor csv y Network File mip_TRAIN_weights v Submit Fig 55 the configuration for the Run use case in the XOR problem DAMEWARE GUI User Manual 38 This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program v My Experiments Workspace mipExp Experiment Status Last Access X Delete gt xorTrain ended 2011 05 30 X 4 xorTest ended 2011 05 30 X Download Addinws File Type Description th a mip_TEST_output csv CSV output and target vector of the test set eS y gt ss mip_TEST log txt log file ch mip_TEST_confusionMatrx txt confusion matrx tt rae MLP_Test_params xmi xmi Experiment Configuration File Fig 56 the output of the TEST use case experiment in the XOR problem At the end of TEST experiment execution the experiment output area should contain a list of output files as shown inFig 54 Also the same file mlp should be selected as Network file input in case you want to execute another training TRAIN FULL cases phase for example when first training session ended in an unsuccessful or insufficient way In this cases the user can execute more training experiments starting learning from the previous on
30. a it dame_ kappa html Photometric Redshifts http dame dsf unina it dame_photoz html Photometric Quasars _http dame dsf unina it dame_qso html Globular Clusters Science Cases http dame dsf unina i dame_gcs html AGN Classification _http dame dsf unina it dame_agn html Sky Transients _http dame dsf unina it dame_td html POE amp Lectures _http dame dsf unina it documents html Science Production me Newsletter http dame dsf unina it newsletters html http dame dsf unina it dameware html notes http dame dsf unina it dameware html faqg YouTube Channel _http www youtube com user DAMEmedia Official website Info http dame dsf unina it Citation Policy _http dame dsf unina it policy Cd Write Us _helpdame gmail com o About Us http dame dsf unina it project_members html Tab 1 Header Area Menu Options Release Notes es gt fO DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program 3 5 Data Management The Data are the heart of the web application data mining amp exploration All its features directly or not are involved within the data manipulation So far a special care has been devoted to features giving the opportunity to upload download edit transform submit create data In the GUI input data i e candidates to be inputs for scientific experiments are basically belonging
31. ad download data files to edit and configure datasets from original data files and or archives e Workspace handling the capacity to create different work spaces depending on the experiment type and data mining model choice Of course in the new release it was impossible to match all important and valid suggestions came from the a release testers In principle not for bad will of developers but mostly because in some cases the requests would needed drastic re engineering of some infrastructure components or simply because they went against our design requirements issued at the very beginning of the project Of course this not implies necessarily that in next releases of the application these requests will not be taken into account Anyway we tried to satisfy as much as possible main requests concerning the improvement of ease to use Also in terms of examples and guided tours in using the available models Don t forget that neophyte users should spend a certain amount of time to read this and other manuals to learn their capabilities and usability topics before to move inside the application This is particularly true in order to understand how to identify the right association of functionality domain and the data mining model to be applied to your own science case But we recall that this is fully reachable by gaining experience with time and through several trial and error sessions 2 Purpose This manual is mainly dedicated to drive users
32. automatically open when user wants to edit data file to create manipulate datasets to upload files or to configure and launch experiments All tabs can be closed by user except the main one Resource Manager Creation of new workspaces When selected and named the new workspace appears in the Workspace List Area Workspace sub window Workspace List Area portion of the main Resource Manager tab dedicated to host all user defined workspaces Upload command When selected the user is able to select a new file to be uploaded into the Workspace Data Area Files Manager sub window The file can be uploaded from external URI or from local user HD Creation of new experiment When selected the user is able to create a new experiment a specific new tab is open to configure and launch the experiment Rename workspace command When selected the user can rename the workspace Delete Workspace command When selected the user can delete the related workspace only if no experiments are present inside otherwise the system alerts to empty the workspace before to erase it 10 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved 10 11 12 13 14 15 16 17 18 19 20 DAME Application User bresciamax gmail com LogOut Ls z L my App Manuals v Model Manuals v Other Services v Science Cases v Documents v Info v DAta Mining amp Exploration
33. daino aa palo 24 0 Mme cdl CO EU OY2ar2 016 MAY VDE NL W271 0G iaia 25 Oe RR RIE 25 SIZE RR IE TEA 31 SII RR RR 31 3 7 1 Re use of already trained networks in niinen 36 TABLE INDEX Tab IH ARMOR na 14 Tab 2 ADDIEViGhONS and QCYONYINS iii 40 Tab 3 ReferenceDocuments iii 41 Tab 4 Apphcable DIO CHINO ria pt 42 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig Fig DAta Mining amp Exploration Program FIGURE INDEX CIAOO RR AO 6 2 The user registration login form to access at the web application iii S AE RA 9 4 An example of e mail received by the user after submission of registration info 9 5 The Web Application starting main page Resource Manager i 10 6 The Web Application main areas and commandS iii Il 7 The right sequence to configure and execute an experiment workflow cccccccceeeeveeeeececeeesaaaeseeseees 12 8 the button New Workspace at left corner of workspace manager window ii 12 9 the form field that app
34. e by resuming the trained weight matrix as input network for future training sessions This operation is the so called resume training phase of a neural network Of course the same XOR problem could be also solved by using another functionality model couple such as Regression_FMLPGA We remind the user to consult when available the related model specific documentation and manuals available from the header menu of the webapp the beta intro web page or the machine learning web page of the official DAME website 39 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved Abbreviations amp Acronyms A amp A Al ANN ARFF ASCII Bok BP BLL CE CSOM CSV DAL DAME DAPL DL DM DMM DMS FITS FL GRID GSOM GUI HW KDD Meaning Artificial Intelligence Artificial Neural Network Attribute Relation File Format American Standard Code for Information Interchange Base of Knowledge Back Propagation Business Logic Layer Cross Entropy Clustering Self Organizing Maps Comma Separated Values Data Access Layer DAta Mining amp Exploration Data Access amp Process Layer Data Layer Data Mining Data Mining Model Data Mining Suite Flexible Image Transport System Frontend Layer FrameW ork Global Resource Information Database Gated Self Organizing Maps Graphical User Interface Hardware Knowledge Discovery in Databases A amp
35. ears after pressing the New Workspace Dutton ccccccccccccccccccccseceeeeeeeeeees 13 10 the active workspace created in the Workspace List Area iii 13 11 The GUI Header Area with all submenus Open iiiii 14 12 The Upload data feature open in a new tab iii 16 13 The Upload data from external URI feature ii 16 14 The Upload data from Hard Disk feature iiiii 17 15 The Uploaded data file in the File Manager sub Window iiii 17 16 The dataset editor tab with the list of available Operations iii 18 17 The Feature Selection operation select columns and put saving name ii 19 18 The Feature Selection operation the new file created ii 19 19 The Column Ordering operation the starting VIeW iii 19 20 The Column Ordering operation new order tO COIUMNS ii 20 21 The Column Ordering operation new file created ii 20 22 The Sort Rows by Column operation step l iii 20 23 The Sort Rows by Column operation step 2 ii 21 24 The Sort Rows by Column operation the new file created ii 21 25 The Column Shuffle operation step l iii 21 26 The Column Shuffle operation the new file created M i Ze 27 The Row Shuf
36. enu button Image Viewer A dedicated tab will appear in the Resource Manager giving the possibility to load and visualize any image previously uploaded or produced in the web application DAME Application User jobfmann gmail com LogOut 2 Model Manuals v Cloud Services v Science Cases v Documents v RESOURCE MANAGER Histogram Scatter Plot 2D Scatter Plot 3D EJ Line Plot Image Viewer Image View V Graphic Options Workspace alfonso y i Image sky fits H Zoom la Load Image kH Save Image as 1 100 Fig 40 The visualization tab As shown in Fig 40 the visualization tab offer the following options e Workspace the user workspace hosting the image Image the name of the image to be visualized Load Image after the selection of workspace and image this button shows the image Crop button used to crop the image Hand button used to move the image Zoom sliding bar used to zoom the image Save Image as button used to save the modified image ADVERTISEMENT multi image fits files are not supported by this functionality 3 7 Experiment Management After creating at least one workspace populating it with input data files of supported type and optionally creatingany dataset file the next logical operation required is the configuration and launch of an experiment In what follows we will explain the experiment configuration and execution by making use of an example very simp
37. er ended state Other data files are created by dataset creation features a list of operations that can be performed by the user starting from an original data file uploaded in a workspace These data files are automatically generated with a special name as output of any of the manipulation dataset operations available Besides these general rules there are some important prescriptions to take care during the preparation of data to be submitted and the setup of any Machine Learning model v Input features to any machine learning model must be scalars not arrays of values or chars In case you could try to find numerical representation of any not scalar or alphanumerical quantities v The input layer of a generic hierarchical neural network must be populated according to the number of physical input features of your table entries There must be a perfect correspondence between number of input nodes and input features columns of your table v All objects rows of an input table must have exactly the same number of columns No rows with variable number of columns are allowed v Hidden layers of any multi layer feed forward model i e layers between input and output ones must contain a decreasing number of nodes usually by following an empirical law given N input nodes the first hidden layer should have 2N 1 nodes at least the optional second hidden layer N 1 and so on v For most of the available neural networks models
38. experiment where starting form an original data file several different files must be prepared and provided to be submitted as input for respectively training test and validate the algorithm chosen for the experiment This pre processing is generally made by applying one or more modification to the original data file for example 17 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program obtained from any astronomical observation run or cosmological simulation The operations available in the web application are the following Fig 16 Feature Selection Columns Ordering Sort Rows by Column Column Shuffle Row Shuffle Split by Rows Dataset Scale Single Column Scale All these operations one by one can be applied starting from a selected data file uploaded in the currently active workspace RESOURCE MANAGER Upload in myfirstWS Workspace E File Editor Workspace my firsts Configuration File opt_and_struct_ceyv gu lt Operation Description t creates a new file containing only the selected Feature Selection i columns It creates a new file containing the new ordering of u i in z ae CECI columns as specified in the column fields t creates a new file containing the sorted rows by Sort Rows by Column Se by specifying the column reference index Column Shuffle It creates a new
39. file containing shuffled columns Row Shuffle t creates a new file containing shuffled rows It creates as many new files as desired each Split by Rows containing the specified percentage of rows The rows distribution can be randomly extracted Use only numerical Dataset Dataset Scale It creates a new file containing all data scaled in the selected range 1 1 or 0 1 It creates a new file containing a single column data scaled in the selected range 1 1 or 0 1 leaving unmodified the other columns in the original order Single Column Scale Fig 16 The dataset editor tab with the list of available operations 3 5 2 1 Feature Selection This dataset operation permits to select and extract arbitrary number of columns contained in the original data file by saving them in a new file of the same type and with the same extension of the original file named as columnSubset_ lt user selected name gt i e with specific prefixcolumnSubset This function is particularly useful to select training columns to be submitted to the algorithm extracted from the whole data file Details of the simple procedure are reported in Fig 17 and Fig 18 18 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved RESOURCE MANAGER Upload in myfirstWS Workspace Workspace myfirstWS File opt_and_struct_csv lt Operation Feature Selection Columns Orde
40. fle operation step l cccscccccccccccccccccccccccccsccnccenssssssseeeeeecccccceessuaaacggaassssseeeeeeseeeeeeeeesaaas 22 28 The Row Shuffle operation the new file CreAted cccccccccccccccceseessseecceeeecaaesssseeceeeeeaaaeeeseeeeeeeeaaaaees 22 29 The Split by Rows operation SD 23 30 The Split by Rows operation the new files created iii 23 31 The Dataset Scale operation Step Liiccscccccccccccccccccccccccccsssccccssssssssseeeeeecccceeesssuuaggasssesseeeeeeeeeeseeeeesaags 24 32 The Single Column Scale operation step l ccccsvissesnceoscuseseeseaoccesncdaadeinaisuctcneebasvesesormbondiaadeiansiseovdies 24 Ihe l storan laD ssciscicescaue su rine E E E E EES 23 AMO 26 ANIA RR RE 27 SO SAM layer scatter D DI OF oerrint r ciano ten 28 VARO DIL RAI 28 O The EMNE TOU anena EE E E A E E a ni 29 39 A m lti layer MED ilaria liana 30 PO TVS ALON GA ie atea 31 41 Creating a new experiment by selecting icon Experiment in the workspace 32 42 The new tab open after creation of a new experiment with the list of available options 32 43 The new state of the experiment configuration tab after the selection of the model 33 44 The configuration options in the Train use CASE csscccccccccccneneesscecccceeeeaeeseseecceeeeaaaaeseeeeeeesaaaaaeeseess 34 45 The configuration options in the Test use Case iii 34 46 The c
41. ile optAndStruct columns colt Float Name It creates a new file containing the new ordering of col2 Float columns as specified in the column fields col3 Float It creates a new file containing the sorted rows by specifying the column reference index col4 Float oon N a WN n It creates a new file containing shuffled columns Float It creates a new file containing shuffled rows col Float It creates as many new files as desired each col7 Float containing the specified percentage of rows The cols Float rows distribution can be randomly extracted Use only numerical Dataset cola Double It creates a new file containing all data scaled in the 10 colo Double selected range 1 1 or 0 1 11 111 Double It creates a new file containing a single column data a scaled in the selected range 1 1 or 0 1 12 col12 Short leaving unmodified the other columns in the original order Fig 25 The Column Shuffle operation step 1 21 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved 3 5 2 5 Row Shuffle v File Manager Workspace imy firstWS B Dow Edit File I amp columnSort_optAndStruct I io columnSubset_dataset 7 r al I opt_and_struct csv tl amp rowShuffle_optAndStruct Gi rowSort_optAndStruct B amp shuffle_optAndStruct Er Type csv csv csv csv csv csv csv DA
42. ing all data scaled in the selected range 1 1 or 0 1 It creates a new file containing a single column data scaled in the selected range 1 1 or 0 1 leaving unmodified the other columns in the original order DAMEWARE GUI User Manual Configuration Column k Column Name 1 ono YO N a WwW N os sb N ae O coli col col3 col4 cols col col7 cols colg col10 col11 col12 Type Float Float Float Float Float Float Float Float Double Double Double Short This dataset operation permits to select an arbitrary order of columns contained in the original data file by saving them in a new file of the same type and with the same extension of the original file named as columnSort_ lt user selected name gt i e with specific prefixcolumnSort Details of the simple procedure are reported inFig 20 Output OK Name Fig 19 The Column Ordering operation the starting view 19 This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program In particular in Fig 20 it is shown the result of several dragging operations operated on some columns By selecting with mouse a column it is possible to drag it in a new desired position At the end the new saved file will contain the new order given to data columns RESOURCE MANAGER File Editor Workspace myfirstWS File opt_and_struct_c
43. into account the training phase inside its operation sequence Apart from the training step a complete scientific workflow always includes a well defined sequence of steps including pre processing or equivalently preparation of data training validation run and in some cases post processing The DMS permits to perform a complete workflow having the following features e A workspace to envelope all input output resources of the workflow e A dataset editor provided with a series of pre processing functionalities to edit and manipulate the raw data uploaded by the user in the active workspace see section 3 5 for details e The possibility to copy output files of an experiment in the workspace to be arranged as input dataset for subsequent execution the output of training phase should become the input for the validate run phase of the same experiment e An experiment setup toolset to select functionality domain and machine learning models to be configured and executed e Functions to visualize graphics and text results from experiment output e A plugin based toolkit to extend DMS functionalities and models with user own applications 3 1 User Registration and Access The DMS makes use embedded to the end user of the Cloud computing infrastructure made by single PCs in combination with GRID resources This requires a reliable level of security in order to launch jobs experiments in a safe and coordinated way This level of securi
44. ity to create and visualize a scatter 3D of any table file previously loaded or produced in the web application There are several options e Workspace the user workspace hosting the table e Table the name of the table to be plotted 28 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program xAxis selection of the x column of table to be plotted yAxis selection of the y column of table to be plotted ZAxis Selection of the z column of table to be plotted Marker_Size size of the marker Color color of the plot bars Marker_Shape shape of the marker Line_Width width of the bars Flip enable the flipping of the x Axis of the plot Flip enable the flipping of the y Axis of the plot Flip enable the flipping of the z Axis of the plot Title title of the plot Xlabel label of the x axis Ylabel label of the y axis Zlabel label of the z axis Grid enable disable the grid in the plot Fog enable disable the fog effect Phi rotation angle in degrees Theta rotation angle in degrees Orientation buttons four predefined couples of Phi and Theta Clear Tab clear the tab Plot creation and visualization of the selected histogram Save Plot As plot saving with user typed name Export in another window the plot will be moved in an independent tab of the web browser Add Tab enable the creation of a mul
45. l 663 pp 752 764 The Fourth Paradigm Microsoft research Redmond Washington USA Artificial Intelligence A modern Approach Second ed Prentice Hall Pattern Classification A Wiley Interscience Publication New York Wiley Neural Networks A comprehensive Foundation Second Edition Prentice Hall A practical applicationof simulated annealing to clustering Pattern Recognition 25 4 401 412 Probabilistic connectionist approaches for the design of good communication codes Proc of the IJCNN Japan Approximations by superpositions of sigmoidal functions Mathematics of Control Signals and Systems 2 303 3 14 no 4 pp 303 314 Program Author Ronald Fisher Bishop C M Bishop C M Svensen M amp Williams C K I Dunham M D Abrusco R et al Hey T et al Russell S Norvig P Duda R O Hart P E Stork D G Haykin S Donald E Brown D E Huntley C L Babu G P Murty M N Cybenko G Tab 3 ReferenceDocuments DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Date 1936 1995 1998 2002 2007 2009 2003 2001 1999 1991 1993 1989 41 ID Al A2 A3 A4 AS A6 A7 AS A9 A10 All AI2 A13 A14 A15 A16 A17 A18 A19 A20 A21 Title Code SuiteDesign VONEURAL PDD NA 0001 Rel2 0 project_plan_ VONEURAL PLA NA 000
46. le not linearly separable XOR problem which can be replicated by the user by using the xor csv and xor_run csv data files downloadable from the beta intro web page http dame dsf unina it dameware html 31 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program The Fig 41 shows the initial step required i e the selection of the icon command nr 7of Fig 6 in order to create the new experiment Model Manuals j Science Cases v RESOURCE MANAGER Workspace v File Manager aaa T l Workspace E New Workspace Uo Plot Editor m Image Viewer MyFirstwS I m hi nr o E n o gt ay i E Dow Edit File Type Last Access SC Dele Rename gt Workspace C Upload cd Experiment Delete B m101 fits fits table 2013 09 02 x f MyFirstWS z uit x B amp optandstruct fits fits table 2013 09 04 x secondWs C x G opt_and struct csv csv 2013 07 01 x New Experiment x Experiment Name MyFirstExp OK Cancel Fig 41 Creating a new experiment by selecting icon Experiment in the workspace Immediately after an automatic new tab appears making available all basic features to select configure and launch the experiment In particular there is the list of couples functionality model to choose for the current experiment The proper choice should be done in order to solve a particular problem I
47. mly extracted Use only numerical Dataset co Double It creates a new file containing all data scaled in the 10 col10 Double selected range 1 1 or 0 1 11 col11 Double It creates a new file containing a single column data scaled in the selected range 1 1 or 0 1 12 col12 Short leaving unmodified the other columns in the original order Fig 23 The Sort Rows by Column operation step 2 v Files Manager Dataset File Type X Delete i Download Last Access i myNewDatasetcolumnSubset fits 2010 06 22 x ih mySecondDatasetcolumnSort fits 2010 06 22 x ii myThirdDatasetrowSort fits 2010 06 22 x ih TRAIN fits x Fig 24 The Sort Rows by Column operation the new file created 3 5 2 4 Column Shuffle This dataset operation permits to operate a random shuffle of the columns contained in the original data file The result is the creation of a new file of the same type and with the same extension of the original file named as shuffle_ lt user selected name gt i e with specific prefixshuffle Details of the simple procedure are reported in Fig 25 and Fig 26 RESOURCE MANAGER File Editor Workspace myfirstWS File opt_and_struct_csv X Operation Feature Selection Columns Ordering Sort Rows by Column Column Shuffle Row Shuffle Split by Rows Dataset Scale Single Column Scale Configuration Deagini Output OK It creates a new file containing only the selected F
48. mmagini a Musica EJ Video jE Computer amp Disco locale C a RECOVERY D HP TONI S FY El m Nome file opt_and_struct Fig 14 The Upload data from Hard Disk feature After the execution of the operation coming back to the main GUI tab the user will found the uploaded file in the Files Manager sub window related with the currently active workspace Fig 15 DAME Application User bresciamax gmail com LogOut Li g me App Manuals v Model Manuals v Other Services v RESOURCE MANAGER Workspace V File Manager Workspace My firstWS _ a a r 7 i Dow amp Edit File Type Last Access M Dele Rename Workspace E Upload jj Experiment 3 Delete A Gi amp opt_and struct csv csv 2013 09 09 primo B x gt G xor csv csv 2013 09 03 I mipqnaExp B cd x gt G amp xor_run csv csv 2013 09 03 x 4 mipexp B E x phat B cd x clashRegr B da x da B da x img1 B da x i V My Experiments f bestarsqna Aa a x Workspace MyfirstwS l 4survey B x Experiment Status Last Access J Delete P MyfirstWS G a x b xorTrain ended 2013 09 03 x b xorTest ended 2013 09 03 x Fig 15 The Uploaded data file in the File Manager sub window 3 5 2 How to Create dataset files If the user has already uploaded any supported data file in the workspace it is possible to select it and to create datasets from it This is a typical pre processing phase in a machine learning based
49. new file names are those filled in by the user in the proper name fields as split1_ lt user selected name gt split2_ lt user selected name gt i e with specific prefixsplit and split2 Details of the simple procedure are reported inFig 29 Fig 30 RESOURCE MANAGER File Editor Workspace myfirstWS Configuration File opt_and_struct_csv 9 lt Operation Description n OK n It creates a new file containing only the selected Feature Selection i 1 coll Float og It creates a new file containing the new ordering of 2 col2 Float 61 n Columns Ordering i ified in th i field columns as specified in the column fields 3 col3 Float It creates a new file containing the sorted rows by i E specifying the column reference index 4 col4 Float File Name optAndStruct Column Shuffle It creates a new file containing shuffled columns 5 cols Float p Row Shuffle It creates a new file containing shuffled rows 6 co Float 70 It creates as many new files as desired each 7 col Float Split by Rows containing the specified percentage of rows The 8 cols Float 1 99 rows distribution can be randomly extracted Use only numerical Dataset 9 col Double Dataset Scale It creates a new file containing all data scaled in the 10 col10 Double 11 colt1 Double File Name optAndStruct t creates a new file containing a single column data scaled in the selected range 1 1 or 0 1 12 col12 Short Pron e n leaving unmodified the
50. ning Suite is the EXPERIMENTS Plotting and Visualizing Fig 1 Suite functional hierarchy By this way as usual in data mining the knowledge discovery process should basically consist of several experiments belonging to a specified functionality domain in order to find the model parameter configuration and dataset parameter space choices that give the best results in terms of performance and reliability The following sections describes in detail the practical use of the DMS from the end user point of view Moreover the DMS has been designed to build and execute a typical complete scientific pipeline DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program hereinafter named workflow making use of machine learning models This specification is crucial to understand the right way to build and configure data mining experiment with DMS In fact machine learning algorithms hereinafter named models need always a pre run stage usually defined as training or learning phase and are basically divided into two categories supervised and unsupervised models depending respectively if they make use of a BoK Base of Knowledge i e couples input target for each datum to perform training or not for more details about the concept of training data see section 3 5 below So far any scientific workflow must take
51. ocessing phase for clustering on text or image files e ESOM Evolving SOM for pre clustering on text or image files Specific related manuals are available to obtain detailed information about the use of the above models see webapp header menu options After the selection of the proper functionality model the tab will show greyed some options and the possibility to select the use case The greyed options like help button will be activated after the selection of the use case to be configured and launched RESOURCE MANAGER File Editor Experiment Setup Q Q Selecta Regression MLP y Select a Running Pun Mode w Functionality Mode Field is Required Fig 43 The new state of the experiment configuration tab after the selection of the model As known data mining models following machine learning paradigm offer a series of use cases see figures below e Train training learning phase in which the model is trained with the user available BoK 33 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program RESOURCE MANAGER File Editor Experiment Setup Select a Select a Running Train iti Functionality Mode Regression_MLP Field is Required Train Set v Validation Set Network File v number of input nodes number of nodes for hidden layer number of output nodes
52. onfiguration options in the Run use Case iii 34 47 The configuration options in the Full use Case ii 35 48 Example of a web page automatically open after the click on the help button 35 49 Some different state of two concurrent experiments i 35 50 An example of Classification _MLP training case for the XOR problem i 36 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program Fig 51 The popup status at the end of the XOR problem experiment iii 36 Fig 52 The list of output files after the XOR problem training experiment 37 Fig 53 The training error scatter plot mlp_TRAIN_errorPlot jpg downloaded from the experiment output list x axis is the training cycle y axis is the training mean square error iii 37 Fig 54 The operation to move the trained network file in the Workspace input file list 38 Fig 55 the configuration for the Run use case in the XOR Problemy iccccccccccccccccccccccsccsssseesecssseeeeeeeeseeeeeenaaas 36 Fig 56 the output of the TEST use case experiment in the XOR problem iii 39 4 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved
53. order Fig 32 The Single Column Scale operation step 1 3 5 3 Download data All data files not only those of supported type listed in the workspace and or in the experiment panels respectively Files Manager and Experiment Manager can be downloaded by the user on his own hard disk by simply selecting the icon labelled with Download in the mentioned panels 3 5 4 Moving data files The virtual separation of user data files between workspace and experiment files located in the respective panels File Manager for workspace files and My Experiments for experiment files is due to the 24 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program different origin of such files and depends on their registration policy into the web application database The data files present in the workspace list File Manager area panel are usually registered as input files 1 e to be submitted as inputs for experiments While others present in the experiment list My Experiments panel are considered as output files i e generated by the web application after the execution of an experiment It is not rare in machine learning complex workflows to re use some output files obtained after training phase as inputs of a test validation phase of the same workflow This is true for example for a
54. ormation of DAME project Board All Rights Reserved
55. ot creation and visualization of the selected histogram Save Plot As plot saving with user typed name Export in another window the plot will be moved in an independent tab of the web browser Add Tab enable the creation of a multi layer line plot as shown in Fig 39 DAME Application User jobfmann gmail com LogOut Model Manuals v Cloud Services v RESOURCE MANAGER Histogram Scatter Plot2D E Scatter Plot 3D Line Plot Graphic View Comparison Among Magnitudes MAG APERI MAG _APER2 MAG ISO N oo MAG APERI MAG APER2 MAG ISO Distributic N 53 Db a N hi o I o 200 400 600 500 1000 1200 1400 1600 1800 2000 7 index v Graphic Options Add Tab Main A B B B Title Comparison Among Magnitu Workspace manual v Color J v XLabel index Table agn_ridotto csv v Line_Width 1 m YLabel MAG_APER1 MAG_APER2 yAxis MAG_ISO v F Flip 7 Grid Clear tab l Piot bey Save Poot as cgil Export in another window Fig 39 A multi layer line plot ADVERTISEMENT whenever the user change any parameter of the current plot it is needed to click the button Plot to refresh the visualized plot 30 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program 3 6 2 Visualization This option can be enabled from the main tab of the GUI by simply clicking on the m
56. r bresciamax gmail com LogOut o By 0 Manuals Model Manuals Documents y RESOURCE MANAGER Upload in primo Workspace Workspace primo Upload URI m from Resource Location http dame dsf unina it docur Upload csv fits table fits image votable png gif ips epg other Fig 13 The Upload data from external URI feature In the second case upload from Hard Disk the Fig 14 shows how to select and upload any supported file in the GUI workspace from the user local HD 16 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program RESOURCE MANAGER File Editor E Upload in myfirstWS Workspace E Workspace myfirstWS _ Hard Disk Tal esv i Apri mes CAI DATASETS data test_1_29Sep2010 n Cerca data_test_1_295ep2010 P Hard Disk Organizza v Nuova cartella d z l Download Nome Ultima modifica Nessun file selezionato 23 Risorse recenti a Upload File gt opt_and_struct 09 2010 12 tab_no_8_9 23 09 2010 13 17 tab_no_8_9_10 23 09 2010 13 19 gt tab_no_8 9 11 23 09 2010 13 19 tab_no_8 10 23 09 2010 13 17 tab_no_8 10 11 23 09 2010 13 18 tab_no_8 11 23 09 2010 13 16 gt tab_no_9_10 23 09 2010 13 17 tab_no_9_10_11 23 09 2010 13 18 tab_no_9 11 23 09 2010 13 16 Raccolte A Documenti I
57. range 1 1 or 0 1 It creates a new file containing a single column data scaled in the selected range 1 1 or 0 1 leaving unmodified the other columns in the original order Configuration Output OK Pe aas 1 coli Float Name 2 col Float 3 col3 Float 4 col4 Float 5 cols Float 6 col Float 7 col7 Float 8 cols Float 9 col9 Double 10 col10 Double 11 col11 Double 12 col12 Short Fig 27 The Row Shuffle operation step 1 v File Manager Workspace my firstwS gt Dow g Edit File columnSort_dataset columnSort_optAndStruct columnSubset_dataset DOL PEFFE opt_and_struct csv Et rowShuffle_optAndStruct a rowSort_optAndStruct n A shuffle_optAndStruct Type csv csv csv csv csv csv csv Last Access 2010 12 01 2010 12 01 x 2010 11 30 x 2010 11 30 x 2010 12 01 X x E 2010 12 01 2010 12 01 Fig 28 The Row Shuffle operation the new file created 22 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program 3 5 2 6 Split by Rows This dataset operation permits to split the original file into two new files containing the selected percentages of rows as indicated by the user The user can move one of the two sliding bars in order to fix the desired percentage The other sliding bar will automatically move in the right percentage position The
58. ring Sort Rows by Column Column Shuffle Row Shuffle Split by Rows Dataset Scale Single Column Scale File Editor Description It creates a new file containing only the selected columns It creates a new file containing the new ordering of columns as specified in the column fields It creates a new file containing the sorted rows by specifying the column reference index It creates a new file containing shuffled columns It creates a new file containing shuffled rows It creates as many new files as desired each containing the specified percentage of rows The rows distribution can be randomly extracted Use only numerical Dataset It creates a new file containing all data scaled in the selected range 1 1 or 0 1 It creates a new file containing a single column data scaled in the selected range 1 1 or 0 1 leaving unmodified the other columns in the original order Configuration O m 1 F 2 VI 3 WI 4 F 5 ra 6 m 7 v 8 A us m 10 m 11 v 12 Column k Column coli col col3 col4 cols cole col7 cols colg col10 col11 col12 Type Float Float Float Float Float Float Float Float Double Double Double Short DAta Mining amp Exploration Program Output Ok Name Fig 17 The Feature Selection operation select columns and put saving name v File Manager Workspace myfirstwWS E Dow amp Edit File Type
59. ription E i mip_TRAIN_errorPlot jpeg jpeg scatter plot of the epochs ws error i mip_TRAIN_error csy csw epoch error file B mip_ TRAIN log txt log file i mip_TRAIN_tmp_weights mip net tmp file i mip TRAIN weights mip trained network file Fig 52 The list of output files after the XOR problem training experiment 0 25f 0 20 0 05 0 00 2 0 10 20 30 40 50 60 70 1 Fig 53 The training error scatter plot mlp_TRAIN_errorPlot jpg downloaded from the experiment output list x axis is the training cycle y axis is the training mean square error The Run case is hence executed to perform scientific experiments on new data Remember also that the input file does not include target values The execution of a Run use case for its nature requires special steps in the DAME Suite These are described in the following As first step we require to have already performed a train case for any experiment obtaining a list of output files train or full use cases already executed In particular in the output list of the train full experiment there is the file m p This file contains the final trained network in terms of final updated weights of neuron layers exactly as resulted at the end of the training phase Depending on the training correctness this file has in practice to be submitted to the network as initial weight file in order to perform test run sessions on input data without target values 37 DAMEWARE G
60. s shown in Fig 34 DAME Application User jobfmann gmail com LogOut p App Manuals v Model Manuals v Cloud Services v Science Cases v l MAG_APERI MAG _APER2 M mac ISO Count 19 5 2060 205 21 0 21 5 22 0 225 23 0 235 240 245 25 0 255 260 265 27 0 27 5 280 285 280 255 30 0 MAG_APER1 MAG_APER2 MAG_ISO Distribution v Graphic Options Add Tab Bin Placement i i Width 0 1 Title Comparison Among Magnitu Main A B B B XLabel MAG_APER1 MAG_APER2 Workspace manual v Bar_Style 2 Steps v YLabel Count Table agn_ridotto csv v Color E lv E Grid xAxis MAG_ISO by E Flip Line_Width 1 v Clear tab Uk Prot Save Plot as cy Export in another window Fig 34 A multi layer histogram plot ADVERTISEMENT whenever the user change any parameter of the current plot it is needed to click the button Plot to refresh the visualized plot 26 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program DAME Application User jobfmann gmail com LogOut amp Model Manuals v Cloud Services v Science Cases v RESOURCE MANAGER Histogram Scatter Plot 2D Scatter Plot 3D E Line Plot Graphic View sScarrer PIOT_21 Main Title Scatter Plot 2D Workspace ppsworkspace v Marker Sze 2 XLabel x Table dataset_train_80 ascii he Color
61. ss of the program official website http dame dsf unina it DAME DOCUMENTATION LIST html 3 GUI Overview following The DMS is a web application accessible through a simple web browser It is structurally organized under the form of working sessions hereinafter named workspaces that the user can create modify and erase You can imagine the entire DMS as a container of services hierarchically structured as in Fig 1 The user can create as many workspaces as desired and populate them with uploaded data files and with experiments created and configured by using the Suite Each workspace is enveloping a list of data files and experiments the latter defined by the combination between a functionality domain and a series one at least of data mining models From these considerations it is obvious that a workspace makes sense if at least one data file is uploaded into So far the first two actions after logged in are respectively to create a new workspace by assigning it a name and to populate it by uploading at least one data file to be used as input for future experiments The data file types allowed by the DMS are reported in the next sections In principle there should be many experiments belonging to a single workspace made by fixing the functional domain and by slightly different variants of a model setup and configuration or by varying the associated models Me philosophy behind the interaction between user and the DMS Data Mi
62. st Access SM Dele a Rename gt Workspace E Upload J Experiment Delete a n Go gt FMLPGA_Train_weights txt ascii 2013 09 03 x f primo B E x FMLPGA_Train_weights txt G xor_run csv E x 4 mipqnaExp B di x _ G xorcsv csv 2013 09 03 x 4 mipexp X uit x f phat G x clashRegr B ii x f tx P E x img1 m d 9 cs a x V My Experiments Workspace bestarsqna C ai x FMLPGAExp Experiment Status Last Access gt Delete 4survey B x E xorTrain ended 2013 09 03 x i P FMLPGAExp B iy x G Download gt Addinws File Type Description al 3 a FMLPGA_Train_trainerrors jpeg jpeg Image file G Er FMLPGA_Train_trainconfusionmat other training output pseudo confusion r a a FMLPGA_Train log ASCI File log iw tar FMLPGA_Train_weights txt ascii trained network weight file EMLDC A Troin narnmoati acral Exnorimont Con fieurrotion Filo Fig 5 The Web Application starting main page Resource Manager 3 2 The command icons The interaction between user and GUI is based on the selection of icons which correspond to basic features available to perform actions Here their description related to the red circles in Fig 6 is reported l di The header menu options When one of the available menus is selected a pop submenu appears with some options Logout button If pressed the GUI and related working session is closed Operation tabs The GUI is organized like a multi tab browser Different tabs are
63. sv dita 0 tio D iptio K Operation Se Column k Column Name Type Output OK Feature Selecti It creates a new file containing only the selected File dataset eature Selection DUMNA 3 col3 Float Name ini i 1 coll Float Columns Ordering It oe anew a Fis pedina of columns as specified in the column 2 col Float It creates a new file containing the sorted rows by Sert Renna by Cona specifying the column reference index 6 col Float Column Shuffle It creates a new file containing shuffled columns 4 cok Float Row Shuffle It creates a new file containing shuffled rows cols Float It creates as many new files as desired each 8 col Float Split by Rows containing the specified percentage of rows The 7 colf Elan rows distribution can be randomly extracted colg Use only numerical Dataset 9 colo Dataset Scale It creates a new file containing all data scaled in the 10 collo Double selected range 1 1 or 0 1 E z 11 col11 Double It creates a new file containing a single column data scaled in the selected range 1 1 or 0 1 12 col12 Short a leaving unmodified the other columns in the original order Fig 20 The Column Ordering operation new order to columns v File Manager Workspace my firstWS E Dow Edit File Type Last Access a A B columnSort_dataset csv 2010 12 01 x columnSort_optAndStruct csv 2010 12 01 xX a I amp columnSubset dataset columnSort_optAndStruct x tl i op
64. t Scale It creates a new file containing all data scaled in the 10 colo Double selected range 1 1 or 0 1 n 11 coli1 Double It creates a new file containing a single column data scaled in the selected range 1 1 or 0 1 12 coli2 Short Single Column Scale Fig 22 The Sort Rows by Column operation step 1 leaving unmodified the other columns in the original order DAMEWARE GUI User Manual 20 This document contains proprietary information of DAME project Board All Rights Reserved RESOURCE MANAGER File Editor Workspace myfirstWS File opt_and_struct_csv Operation Feature Selection Columns Ordering Sort Rows by Column Column Shuffle Row Shuffle Split by Rows Dataset Scale Single Column Scale DAta Mining amp Exploration Program Configuration Denesiotion Select 3 m OK It creates a new file containing only the selected Index columns colt Fost Output It creates a new file containing the new ordering of col2 Float File columns as specified in the column fields Name col3 Float It creates a new file containing the sorted rows by specifying the column reference index com Float oo AD N WIN n It creates a new file containing shuffled columns Float It creates a new file containing shuffled rows co Float It creates as many new files as desired each col Float containing the specified percentage of rows The cols Float rows distribution can be rando
65. t depends basically on the dataset to be used as input and on the output the user wants to obtain Please refer to the particular model reference manual for more details RESOURCE MANAGER Experiment Setup Workspace MyFirst VS Experiment MyFirstExp Select a Functionality _______ amp m _ __CI Classification _FMLPGA Classification_MLP Classification_MLPQNA Classification_SVM Clustering_CSOM Clustering _E SOM Clustering GSOM Clustering_Kmeans Clustering_SOM Clustering_SOM_Auto Clustering SOM_Kmeans Clustering SOM_TWL Clustering_SOM_UmatCC Feature_Extraction_PPS Regression_FMLPGA Regression_MLP Regression_MLPQNA Regression_SVM Fig 42 The new tab open after creation of a new experiment with the list of available options The user can choose between classification regression or clustering type of functionality to be applied to his problem Each of these functionalities can be achieved by associating a particular data mining model chosen between following types e MLP Multilayer Perceptron neural network trained by standard Back Propagation descent gradient of the error learning rule Associated functionalities are classification and regression 32 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program e KFMLPGA Fast Multilayer Perceptron neural network trained by Genetic Algorithm
66. t_and_struct csv csv 2010 11 30 x Fig 21 The Column Ordering operation new file created 3 5 2 3 Sort Rows by Column This dataset operation permits to select an arbitrary column between those contained in the original data file as sorting reference index for the ordering of all file rows The result is the creation of a new file of the same type and with the same extension of the original file named as rowSort_ lt user selected name gt i e with specific prefixrowSort Details of the simple procedure are reported in Fig 22 Fig 23 and Fig 24 RESOURCE MANAGER File Editor Workspace myfirstWS File opt_and_struct_csv lt Operation Description Select OK ini Index a It creates a new file containing only the selected 1 Float columns Output Cites Orde tes It creates a new file containing the new ordering of 2 col Float File g columns as specified in the column fields 3 a Float Name 3 4 It creates a new file containing the sorted rows by e specifying the column reference index 4 soni Float i Column Shuffle It creates a new file containing shuffled columns 5 cols Float 7 Row Shuffle It creates a new file containing shuffled rows 6 cok Float 8 It creates as many new files as desired each 7T col Float 9 Split by Rows containing the specified percentage of rows The Ae Float 10 rows distribution can be randomly extracted 11 Use only numerical Dataset 9 co Double 12 Datase
67. ta Mining amp Exploration Program Last Access X Dele 2010 12 01 x la 2010 11 30 x 2010 11 30 x 2010 12 01 x 2010 12 01 X 2010 12 01 x 2010 11 30 x t Fig 26 The Column Shuffle operation the new file created This dataset operation permits to operate a random shuffle of the rows contained in the original data file The result is the creation of a new file of the same type and with the same extension of the original file named as rowShuffle_ lt user selected name gt i e with specific prefixrowShuffle Details of the simple procedure are reported in Fig 27 and Fig 28 RESOURCE MANAGER File Editor Workspace myfirstWS File opt_and_struct_csv Operation Feature Selection Columns Ordering Sort Rows by Column Column Shuffle Row Shuffle Split by Rows Dataset Scale Single Column Scale Description It creates a new file containing only the selected columns It creates a new file containing the new ordering of columns as specified in the column fields It creates a new file containing the sorted rows by specifying the column reference index It creates a new file containing shuffled columns It creates a new file containing shuffled rows It creates as many new files as desired each containing the specified percentage of rows The rows distribution can be randomly extracted Use only numerical Dataset It creates a new file containing all data scaled in the selected
68. ted in the File Manager A file present in both areas can be used as input either as output in the experiments Plot Editor When pressed open in the resource manager four tabs histogram scatter 2d scatter 3d and line plot each tab is dedicated to a specific type of plot Image Viewer When pressed open a new tab in the resource manager dedicated to the visualization of an image The image file 1s intended already loaded in the File Manager RESOURCE MANAGER 3 Workspace v File manager Q Workspace 2 New vorkspaci ilo Plot Editor 19 tal Image Viewer 20 FMLPGAExp E Dow Edt File Type Last Access XM Dele a Rename Workspace gt Upload cd Experiment Delete I IZ 11 Gi 1 Prca rran_weignts ve asci 2013 09 03 13x 8 4 primo 6 B 7 da 9 x FMLPGA_Train_weights txt 3 3 amp xor_run csv ___ x P mipqnaExp B a x G xor csv csv 2013 09 03 x f mipexp B x f clashRegr B da x f tx C a x img1 f mg e a x V My Experiments 14 z Workspace P bestarsqna B cd x FMLPGAExp Experiment 15 Status Last Access MX Delete 4survey eri a xorTrain ended 2013 09 03 16 x FMLPGAExp GB cd x 2 Download AddinwS File Type Description 17 Go 18 a FMLPGA_Train_trainerrors jpeg jpeg Image file 5 ar FMLPGA_Train_trainconfusionmat other training output pseudo confusion r Gi a FMLPGA_Train log ASCH File log iw a FMLPGA_Train_weights txt asci trained network weight file D EMILOC A Tesio
69. ti layer scatter plot 3D ADVERTISEMENT whenever the user change any parameter of the current plot it is needed to click the button Plot to refresh the visualized plot Model Manuals v Cloud Services v RESOURCE MANAGER Histogram J Scatter Plot 2D Scatter Plot 3D DI Line Plot Graphic View Line_Plot o 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 v Graphic Options Add Tab ai Title Line Plot Workspace WS20111221 y Culi E m XLabel X Table dataset_test_20 ascii hy Line_Width 1 YLabel Y yAxis zspec ly E Flip Grid Clear tab Iih Prot bed Save Piot as fl Export in another window Fig 38 The Line Plot tab 29 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program As shown in Fig 38 there is the possibility to create and visualize an histogram of any table file previously loaded or produced in the web application There are several options e Workspace the user workspace hosting the table Table the name of the table to be plotted yAxis selection of the column of table to be plotted Color color of the plot line Line_Width width of the line Flip enable the flipping of the x Axis of the plot Title title of the plot Xlabel label of the x axis Ylabel label of the y axis Grid enable disable the grid in the plot Clear Tab clear the tab Pl
70. ty is obtained by an accounting procedure that foresees an initial registration for new users in order to activate their account on the DAME Suite After activation all subsequent accesses will require login and password as defined by the user at the registration stage The user registration login entry page is shown in Fig 2 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program DAME DAta Mining amp Exploration is an innovative Tore sid general purpose Web based distributed data mining I Saas infrastructure specialized in Massive Data Sets exploration with machine learning methods New on DAME WebApp For news documentation and FAQ information please click i 5 You can obtain the access by following Technical Support a simple registration procedure helpdame AT gmail com 1 Compile the registration form click Register Now button Skype helpdesk 2 Immediately after you will receive by e mail a welcome message 3 Check for an e mail message with your account confirmation 4 Go back at this page and sign in m cd A INAF OACN B Fi Fig 2 The user registration login form to access at the web application New users must be registered by following a very simple procedure requiring to select Register Now button on that page The registration form requires the following information to be filled
71. with exceptions of Random Forest and SVM the class target column should be encoded with a binary representation of the class label For 15 DAMEWARE GUI User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program example if you have 3 different classes you must create three different columns of targets by encoding the 3 classes as respectively 001 010 100 Confused Well don t panic please Let s read carefully next sections 3 5 1 Upload user data As mentioned before after the creation of at least one workspace the user would like to populate the workspace with data to be submitted as input for experiments Remember that in this section we are dealing with supported data formats only DAME Application User bresciamax gmail com LogOut Li By 00 Manuals Other Services v RESOURCE MANAGER Upload in primo Workspace Workspace primo Upload scure k from Fig 12 The Upload data feature open in a new tab As shown in Fig 12 when the user selects the upload command label nr 6 in the Fig 6 a new tab appears The user can choose to upload his own data file from respectively from any remote URI a priori known or from his local Hard Disk In the first case upload from URI the Fig 13 shows how to upload a supported type file from a remote address DAME Application Use

Download Pdf Manuals

Related Search

Related Contents

CMP-4SSC-S - 電子看板Comabo    Acronis SharePoint Explorer  LG Electronics RCT689H DVD Recorder User Manual  Mode d`emploi ADP 5300  Télécharger  un regard de psychanalyste sur la guerre économique  curso de experto en mantenimiento industrial teoría y prácticas de  Whitehaus Collection WH1-114RTB-WH Installation Guide  

Copyright © All rights reserved. Failed to retrieve file