Knowledge Extraction Evolutionary Learning Reference manual
Contents
1. The build xml changes dynamically with any new version of KEEL thus its is not possible to fully describe its structure here However it is possible to describe which part of the file must be changed to allow the inclusion of new methods Firstly the jar target must be found It should have the following struc ture lt target name jar depends compile description Build jars gt The jar target is composite by a great number of tasks every one dealing with the construction of a jar file for each method Inside this target the construction of the new jar file must be described as another task Here is a valid example lt jar jarfile distMet KNN jar manifest src keel Algorithms Lazy Learning KNN Manifest gt lt fileset dir bin includes keel Algorithms Lazy_Learning KNN class gt lt fileset dir bin includes keel Algorithms Preprocess Basic class org core x x x class keel Dataset class keel Algorithms Lazy_Learning x class gt lt jar gt The task must define the locations of the new jar file and their corre sponding manifest file Also must include the files from the classes which compose the method Also the files from the imported classes are required to fully describing the task KEEL Reference Manual Page 9 of 31 Method Description files 2 Method Description files Every method in KEEL e g a preproc2. Knowledge Extraction based on Evolutionary Learning Reference manual Version 1 0 Date 20 5 2009 CONTENTS CONTENTS Contents 1 Basic KEEL developement guidelines 3 ET Introduction irie ei a ee Re a a Se r aa 3 1 2 Developing a new method a 3 121 Reading of the configuration file 4 1 2 2 Developement of the method 5 1 2 3 Writing the output files su orcos ae a en 5 1 2 4 Registering the method in KEEL 5 1 2 5 Making the use case files vii ba Ge ee 8 1 2 6 Building the executables 2as 5 ai 9 2 Method Description files 10 211 Header Sea In eke oe ace a Na aa IE So 10 22 Parameters or ike este a EP RA AN en es 6 11 202 Exampleof use Ae ANU hohe e a 12 3 Method Configuration files 13 Oil PUPS Esc E Wot ent ae aut he 13 92 NOW EMS vt le ae eo BE EE Wee ONE ee Sue ae 14 Bio Parametrs i eee ce ee ae ee ae Gee ae Pe ea 14 34 Example ofuse 22 st es Se isle ad 14 4 Data files 15 Al Header aa anang Sent A Dn AR 15 ado Data na E E ik oh BB ees Be Ban 16 4 3 Example of use srt La a Seiko done oo DA eS 17 5 Output files 18 Sal Example o use Slee Ban a to eek Stace eae Ge tect oss 18 6 Use Case files 20 61 Name a io ne Bede A wa En o NE ta 20 KEEL Reference Manual Page 1 of 31 CONTENTS CONTENTS 6 2 7RETERENCES gt Na AN ee aa Be et he cat dt ts Bae na 21 6 3 General description oss Coe e In Poe i AA 21 64 Example erns gues es Muell
3. The API Dataset allows the presence of missing values in the data files defined with the lt null gt or tokens However only input attributes can present missing values If a missing value is detected in an output attribute a Output ValueNotKnownException will be cast aborting the processing of the data file 7 2 4 Train and test files The semantic verifications performed by the API Dataset will vary depend ing on the concrete data file processed Concretely the actions performed are e The definition of the attributes is taken from the training file e During the test file reading the definitions of the attributes are checked If they are not consistent with the ones read from the training file the processing of the test file is aborted Moreover the inputs and outputs defined by the test file must be the same which were defined by the training file Otherwise the processing of the test file will be aborted KEEL Reference Manual Page 27 of 31 API Dataset Description of the classes 7 3 Description of the classes The API Dataset is composed by four main classes e InstanceSet This class contains a complete set of instances defining a data base e Instance This class represents a single instance e Attributes This static class contains definitions about every attribute of the data contained in the Instance set e Attribute This class contains relevant information about a single attribute The next sub
4. continuous gt Yes lt continuous gt lt integer gt Yes lt integer gt lt nominal gt Yes lt nominal gt lt missing gt Yes lt missing gt lt imprecise gt No lt imprecise gt lt multiclass gt Yes lt multiclass gt lt multioutput gt No lt multioutput gt Continuous The method is able to work with continuous values Integer The method is able to work with integer values Nominal The method is able to work with nominal values Missing The method is able to handle missing values Imprecise Value The method is able to work with imprecise values Multiclass The method is able to work with problem which defines more than 2 classes Multioutput The method is able to work with data which defines more than 1 output for each instance KEEL Reference Manual Page 7 of 31 Basic KEEL developement guidelines Developing a new method When the header input and output sections were completely defined then the new registry can be place inside the corresponding master description file Below is shown a valid example of registry lt method gt lt name gt Disc UniformWidth name dfamily Discretizers family cjar file Disc UniformWidth jar jar file lt problem_type gt unspecified lt problem_type gt lt input gt lt continuous gt Yes lt continuous gt lt integer gt Yes lt integer gt lt nominal gt Yes lt nominal gt lt missing gt Yes lt missing gt lt imprecise gt No lt imprecise gt lt multic
5. instance only the positions with INTEGER or REAL attributes values will produce a value e getOutputNominal Values Returns an array containing all the output values of the instance only the positions with NOMINAL attributes values will produce a value e getOutputMissing Values Returns a boolean array defining which output values are missing e getOutputReal Value Returns the value of a concrete output attribute only the positions with INTEGER or REAL attributes values will produce a value e getOutputNominalValue Returns the value of a concrete output attribute only the positions with NOMINAL attributes values will produce a value e getInputMissing Value Returns a boolean value defining if the out put value is missing e getAllInputValues Returns an array containing all the input values REAL values are returned as double values INTEGER values are casted to double NOMINAL values are transformed to INTEGER and casted to double e getAllOutputValues Returns an array containing all the output val ues REAL values are returned as double values INTEGER values are casted to double NOMINAL values are transformed to INTEGER and casted to double KEEL Reference Manual Page 29 of 31 API Dataset Description of the classes 7 3 3 Attributes Attributes is an static class which stores the definitions of the attributes represented in the data set It contains an array of Attribute objects and two additional array
6. random number generator if the method needs it Also the names and paths of the input and output files are specified inside Usually a KEEL method employs two input data files The training file containing the data set which should be employed in the train phase of the method and the test file containing the data set which should be employed in the test phase In addition any method excepting the preprocessing methods and the test methods specify a third file the validation file This file contains a copy of the original dataset of the experiment which can be used in comparisons with the train data The format of the data files is explained in section 4 These files must be handled with care because they will be employed not only by the method KEEL Reference Manual Page 4 of 31 Basic KEEL developement guidelines Developing a new method but also by the KEEL API dataset see section 7 in order to load and check the data in an efficient way Any KEEL method must define at least two output files A train output file and a test output file In addition it is possible to define additional output files in the configuration file They will be explained in the next subsections of this guide 1 2 2 Developement of the method The development of the method can be done in any programming environ ment The only requirements are The method must be developed with the Java programming language and it must employ a package structure w
7. to work with discretized values Integer The method is able to work with integer values Nominal The method is able to work with nominal values Value Less The method is able to handle missing values Imprecise Value The method is able to work with imprecise values KEEL Reference Manual Page 21 of 31 Use Case files Example lt generalDescription gt lt type gt General type of method lt type gt lt objective gt Objective of the method lt objective gt lt howWork gt Explanation of how works lt howWork gt lt parameterSpec gt lt param gt Parameter one lt param gt lt param gt Parameter two lt param gt lt parameterSpec gt lt properties gt lt continuous gt Yes lt continuous gt lt discretized gt Yes lt discretized gt lt integer gt Yes lt integer gt lt nominal gt Yes lt nominal gt lt valueLess gt Yes lt valueLess gt lt impreciseValue gt Yes lt impreciseValue gt dl properties lt generalDescription gt 6 4 Example The last part of the use case is employed to show an example of utilization of the method It can be employ any number of lines though it is not recommended to place huge examples here lt example gt A example of utilization of the method lt example gt 6 5 Example of use A valid example of a use case file is shown in the next page KEEL Reference Manual Page 22 of 31 Use Case files Example of use lt method gt lt name gt
8. Prototype Nearest Neighbor lt name gt lt reference gt lt ref gt Chin Liang Chang Finding Prototypes for Nearest Neighbor Classifiers IEEE Trans on Computers vol c 23 No 11 1179 1184 lt ref gt reference lt generalDescription gt lt type gt Preprocess Method lt type gt lt objective gt Reduce the size of the training set without losing precision or accuracy in order to a posterior classification lt objective gt lt howWork gt This algorithm merge nearest prototypes of the set if classification accuracy of the original training data set does not decrease and if they have got the same class If not they are removed from the set lt howWork gt lt parameterSpec gt lt param gt Percentage of prototypes Real lt param gt lt parameterSpec gt lt properties gt lt continuous gt Yes lt continuous gt lt discretized gt Yes lt discretized gt lt integer gt Yes lt integer gt lt nominal gt Yes lt nominal gt lt valueLess gt No lt valueLess gt lt imprecise Value gt No lt impreciseValue gt dl properties lt generalDescription gt lt example gt Problem type Classification Method PG PNN Dataset iris Parameters default values We can see output set in Experimentl Results PGPNN lt example gt lt method gt KEEL Reference Manual Page 23 of 31 API Dataset 7 API Dataset One of the main components of KEEL is the API Dataset It manages the entire process of acquisit
9. ame Reference General Description Example lt method gt Name The name of the method Reference A list of references associated with the method General Description Generic information about the method Example A example about the use of the method 6 1 Name The first part of the use case contains the name of the method enclosed by lt name gt tags lt name gt Name of the method lt name gt KEEL Reference Manual Page 20 of 31 Use Case files Reference 6 2 Reference The second part of the use case contains a list of associated references They are enclosed each one by lt ref gt tags lt reference gt lt ref gt First reference lt ref gt lt ref gt Second reference lt ref gt lt reference gt 6 3 General description The general description describes some common features about the method as its objective parameters type of data which can be handle etc It is composed by the following fields Type General type of method Objetive Objective of the method How work A brief explanation about how the method works Parameter spec A specification of each parameter of the method They are enclosed each one by lt param gt tags Properties Generic properties of the methods Each field can contain yes or no strings defining the following capabilities of the method Continuous The method is able to work with continuous values Discretized The method is able
10. ch is fairly similar to WEKA arff format Each KEEL data file is composed by 2 sections Header Basic metadata describing the data set Data Content of the dataset In both sections it is possible to insert comments by employing the Yo character 4 1 Header The header is composed by the following metadata relation bupa2 attribute mcv nominal a b c attribute alkphos integer 23 138 attribute sgpt integer 4 155 attribute sgot integer 5 82 attribute gammagt integer 5 297 attribute drinks real 0 0 20 0 attribute selector true false inputs mcv alkphos sgpt sgot gammagt drinks outputs selector relation The name of the data set attribute Describes one attribute of the data a column It is possible to define three different types of attributes KEEL Reference Manual Page 15 of 31 Data files Data integer attribute lt name gt integer min max real attribute lt name gt real min max nominal attribute lt name gt Valuel value2 valueN The lt name gt is the identifier of the attribute Its maximum length allowed is 12 characters The min and max values fon integer and real attributes and the list of possible values for nominal attributes are optional If they are missing the corresponding values will be extracted from the data by the KEEL data process module inputs Identifiers of the attributes which must be processed as
11. describe its structure because any KEEL method must be able to completely read it in order to get the values of its parameters specified in each execution Each configuration file has the following structure algorithm Name of the method inputData A list with the input data files of the method outputData A list with the output data files of the method parameters A list of parameters of the method containing the name of each parameter and its value one line is employed to each one 31 Input files The files in the inputData list must be separated by one space being each one surrounded by quotation marks If a validation file is employed by the method the files will appear in the following order input data lt training file gt lt validation file gt lt test file gt If not the order employed will be input data lt training file gt lt test file gt The validation file is a copy of the original train data employed at the start of the experiment It is often employed for comparison tasks between the initial data and training data when it has been preprocessed KEEL Reference Manual Page 13 of 31 Method Configuration files Output files 3 2 Output files e A file with the output for training data e A file with the output for test data e Optional Additional output files Additional output files can be specified in this list Although the will be not managed by other KEEL methods it is possibl
12. e pig dc e os a 22 6 5 Example of use booed swe aL MR eS 22 7 API Dataset 24 7 1 Data files grammar persa tle SR eee hn Gas 24 7 2 Semantic restrictions of the datafiles 26 FQ Attributes 22 weve aoe Ren ma a 26 7 2 2 Inputs and outputs definition 26 7 2 3 Missing values he hee Be See A 27 7 2 4 Train and test files 27 7 3 Description of the classes dt e De e 28 73 1 InistanceSet gt 2 beb a a a 28 7312 rl A korg Rae Ay hy etre ty bb deh eGo E 28 7 33 Attributes a wig Sae Ten naa oe ae E 30 DIA Atribute vse bog a See es hs ele Bl A Ee 30 KEEL Reference Manual Page 2 of 31 Basic KEEL developement guidelines 1 Basic KEEL developement guidelines 11 Introduction The purpose of this document is to describe some basic concepts about the structure of KEEL Knowledge Extraction based on Evolutionary Learning and the format of its internal files The aim of this section is to present the KEEL framework describing some guidelines to help a potential developer to build new methods inside the KEEL environment The next sections will deal with the formats of the configuration files of KEEL including data sets files method descriptions and so on Finally the last section describes the API dataset of KEEL which is used to handle and check the data sets files 1 2 Developing a new method Before to start the task of developing a new method inside of KEEL envi ronment
13. e to employ them to extract useful information from the execution of the method representation of models extracted additional outputs performance measures They must be marked with the extension txt 3 3 Parameters The rest of the configuration file is used to describe the values of the param eters In each line appears one parameter followed by a sign and its assigned value If the method needs a seed to initialize a random number generator it must be the first parameter described employing seed as name of the parameter 3 4 Example of use This is a valid example of a Method Configuration file data files lists are not fully shown algorithm genetic algorithm inputData dataset iris iris11 dat outputData result ga iris resultl tra seed 1234578 nGenerations 500 cross two points crossProbability 0 6 mutationProbability 0 2 KEEL Reference Manual Page 14 of 31 Data files 4 Data files In KEEL the data sets are managed by plain ASCII text files with the dat extension Usually they are located under the dist data directory each one in its own folder which also should contains the partitions created from the whole data set In addition preprocess methods will also create data files as its output which will be placed on the datasets directory of its experiment This section describes the format employed to define them whi
14. ess method a test has assigned a XML file which describes its main characteristics This file will be employed by the KEEL GUI to allow the user to select the values of the parameters of any execution of the method The KEEL Method Description files are located under the dist algorithm directory inside of the folder where its associated JAR file is generated e g dist algorithm methods Each Method Description file is an XML composed by a unique root node lt algorithm_specification gt This node is divided into two parts lt algorithm_specification gt Header Parameters lt algorithm_specification gt Header Basic information about the method Parameters A list of parameters of the method 2 1 Header The header is composed by four nodes lt name gt K Nearest Neighbors Classifier lt name gt lt nParameters gt 2 lt nParameters gt lt seed gt 0 lt seed gt lt nOutput gt 1 lt nOutput gt lt Name gt The name of the method lt nParameters gt The number of parameters of the methods must be 0 or highger Seed values employed to initialize random number generators are not counted here lt Seed gt Defines if the method will need a seed to initialize a random number generator Valid values are 1 if a seed is needed or 0 if not KEEL Reference Manual Page 10 of 31 Method Description files Parameters lt nOutput gt The number of additional output files which wil
15. file have been found please ask to a KEEL project manager if it is not clear what file have to be modified a new registry containing the definition of the method must be created The KEEL master description file registers have the following structure lt method gt Header Input Output lt method gt The header is composed by four nodes lt name gt Disc UniformWidth lt name gt lt family gt Discretizers lt family gt lt jar_file gt Disc UniformWidth jar lt jar_file gt lt problem_type gt unspecified lt problem_type gt Name The name of the method Family The category of the method Jar File The name of the Jar file which contains the method KEEL Reference Manual Page 6 of 31 Basic KEEL developement guidelines Developing a new method Problem Type The class of problems which can manage the method There are defined 4 classes e Classification for supervised classification problems e Regression for regression problems e Unsupervised for unsupervised classification problems e g clus tering e Unspecified for any problem supervised classification unsu pervised classification or regression The input and output parts defines the types of data which the method is able to manage both in input data and output data Their fields must specify which types are allowed by employing yes and no keywords A description of the fields is shown as follows lt
16. ghest value of the parameter valid only in integer and real parameters lt item gt A text value for the parameter it can be employed only in list parameters lt Default gt Default value of the parameter 2 3 Example of use This is a valid example of a Method Description file lt algorithm_specification gt lt name gt K Nearest Neighbors Classifier lt name gt lt nParameters gt 2 lt nParameters gt lt seed gt 0 lt seed gt lt nOutput gt 1 lt nOutput gt lt parameter gt lt name gt K Value lt name gt lt type gt integer lt type gt lt domain gt lt lowerB gt 1 lt lowerB gt lt upperB gt 100 lt upperB gt lt domain lt default gt 1 lt default gt lt parameter lt parameter gt lt name gt Distance Function lt name gt lt type gt list lt type gt lt domain gt lt item gt Manhattan lt item gt lt item gt Euclidean lt item gt lt domain gt lt default gt Euclidean lt default gt lt parameter dlalgorithm specification KEEL Reference Manual Page 12 of 31 Method Configuration files 3 Method Configuration files In KEEL every method uses a configuration file to extract the values of the parameters which will be employed during its execution Although it is generated automatically by the KEEL GUI by using the information contained in the corresponding method description file and the values of the parameters specified by the user it is important to fully
17. hanged by its nearest correct value i e if the value is greater than the maximum it is re placed by the maximum if it is lower than the minimum it is replaced by this one e Nominal attributes The new value is accepted and the domain of the attribute is enlarged adding the new value In addition the flag newValuelnTest is marked on Finally if one of these cases appears the API Dataset throws a Test DataBoundsExcedeedException to inform about the changes performed How ever the files will be parsed correctly 7 2 2 Inputs and outputs definition The definition of inputs and outputs in the data files is optional The API Dataset will automatically extract the missing definitions following these rules e If no outputs are defined If no input are defined the last attribute is taken as output The remaining ones will be taken as inputs KEEL Reference Manual Page 26 of 31 API Dataset Semantic restrictions of the data files If there are some inputs defined the attributes not marked as inputs will be taken as outputs e If no inputs are defined the attributes not marked as outputs will be taken as inputs e If inputs and outputs are defined those attributes who are not cur rently defined in one of these categories are discarded Also it is important to note that the inputs and outputs attributes will be defined in the same order as they appear in the header of the data file 7 2 3 Missing values
18. hose root will be the keel src directory where the sources of any KEEL method are located 1 2 3 Writing the output files As is explained before at least two output files must be produced by the method the train output file and the test output file Its format is described in section 5 If it is desired to employ additional output files they also can be created at the end of the execution on the method These additional files will get its name from the configuration file Also it is important to note that in order to let the KEEL GUI automatically generate the names of these files the number of additional outputs of the methods must be placed in the corresponding method description file 1 2 4 Registering the method in KEEL When the method have been fully coded it must be registered in the KEEL configuration files to allow the KEEL GUI to employ the new method The first step is to create a method description file The format of these files is fully described in section 2 The second step involves modifying the master description file of each category method Currently 11 categories are defined e Discretization KEEL Reference Manual Page 5 of 31 Basic KEEL developement guidelines Developing a new method e Educational Methods e Educational Preprocess e Feature Selection e Instance Selection e Method e Postprocess e Preprocess e Tests e TransOthers e Visualize When the correct master description
19. inputs outputs Identifiers of the attributes which must be processed as outputs The inputs and outputs definitions are optional If they are missing all the attributes will be considered as input attributes except the last which will be considered as output attribute 4 2 Data The data instances are represented as rows of comma separated values where each value corresponds to one attribute in the order defined by the header Missing or null values are defined as null or If the dataset corresponds to a classification problem the output type must be nominal attribute selector true false outputs selector data a 92 45 27 Bil O 0 true a 64 59 32 23 lt null gt false b 54 lt null gt 16 54 0 0 false KEEL Reference Manual Page 16 of 31 Data files Example of use If the dataset corresponds to a regression problem the output type must be real attribute selector real 0 0 20 0 outputs selector data a 92 45 27 Sl 00 0 9 ay ot 59 32 23 lt null gt 17 5 b 5 nul l6 5 00 35 4 3 Example of use This is a valid example of a data file relation bupa2 attribute mcv nominal a b c attribute alkphos integer 23 138 attribute sgpt integer 4 155 attribute sgot integer 5 82 attribute gammagt integer 5 297 attribute drinks real 0 0 20 0 attribute selector true false inputs mcv alkphos sgpt sgot ga
20. ion processing and validation of the data files offering the data sets to the developer in a suitable way freeing him from the task of acquiring the data needed to perform any experiment This section describes three key concepts of the API Dataset e Data files grammar The grammar employed to define the data files Any file generated by this grammar will be a valid data file according to the rules shown in section 4 e Semantic restrictions of the data files Apart from the syntax restric tions some semantic verifications are performed by the API Dataset over the data files e Description of the classes To close this section the main public classes of the API Dataset are described 7 1 Data files grammar In this subsection is shown the grammar which describes the format of the KEEL data files The final tokens of the grammar are e Denotes the void production It is also known as A or e e IDENT Denotes an identifier IDENT A Z a z 0 9 e INTEGER Is an integer value INTEGER 0 9 e REAL Is a real value REAL 0 9 0 9 principal gt Relation gt Attributes gt Inputs gt Outputs gt Data Relation gt relation IDENT KEEL Reference Manual Page 24 of 31 API Dataset Data files grammar Attributes gt attribute IDENT attributeType Attributes gt attributeType gt integer intBoundaries gt real realBoundaries gt IDENT idL
21. ist intBoundaries gt INTEGER INTEGER gt realBoundaries gt REAL REAL gt idList gt IDENT idList gt Inputs gt inputs IDENT idList gt Outputs gt outputs IDENT idList gt Data gt data dataList dataList gt lineData dataList gt lineData gt IDENT lineDataCont lineDataCont gt IDENT lineDataCont2 lineDataCont2 gt IDENT lineDataCont2 gt KEEL Reference Manual Page 25 of 31 API Dataset Semantic restrictions of the data files 7 2 Semantic restrictions of the data files 7 2 1 Attributes An attribute can be defined as integer real or nominal as the grammar of the data files defines It is optional to define de minimum and maximum values or the list of values for any attribute if they are not defined they correct values will be extracted during the processing of the training file Anyway if they are defined for integer or real attributes the minimum value defined must be lower than the maximum This way the limits of the values for any attribute will be established during the processing of the training file However it is possible to find values in the test file which exceed the limits for a concrete attribute i e in some schemes of cross validation Depending of the type of the attribute the actions performed by the API dataset are the following e Integer or Real attributes The new value is c
22. l be gener ated by the method 2 2 Parameters The parameters of the method are listed consecutively A lt parameter gt node is employed to describe each one Each lt parameter gt is composed by the following nodes lt parameter gt lt name gt K Value lt name gt lt type gt integer lt type gt lt domain gt lt lowerB gt 1 lt lowerB gt lt upperB gt 100 lt upperB gt lt domain lt default gt 1 lt default gt lt parameter lt Name gt The name of the parameter lt Type gt Type of parameter KEEL defines four valid types integer An integer value Can be positive 0 or negative enw real A real value The dot is employed as decimal separator text A string of text list A predefined list of text options When employing text parameters no checking operations are done by the KEEL GUI Thus the use of list parameters is recommended when a fixed number of text options are defined so the method does not have to check the parameters by itself lt Domain gt The domain of the parameter For list parameters is manda tory For text parameters cannot be defined For integer and real parameters is optional if it is not defined the KEEL GUI will not check its value lt lowerB gt The lower value of the parameter valid only in integer and real parameters KEEL Reference Manual Page 11 of 31 Method Description files Example of use lt upperB gt The hi
23. lass gt Yes lt multiclass gt lt multioutput gt No lt multioutput gt lt input gt lt output gt lt continuous gt No lt continuous lt integer gt No lt integer gt lt nominal gt Yes lt nominal gt lt missing gt Yes lt missing gt lt imprecise gt No lt imprecise gt lt multiclass gt Yes lt multiclass gt lt multioutput gt No lt multioutput gt lt output lt method gt 1 2 5 Making the use case files When developing a new method it is important to document properly its functions and objectives Also the users should be able to look up relevant information about the method a brief description some references the description of its parameters etc when they select the method in KEEL To manage this information the KEEL GUI defines the use case files which are XML files containing all the relevant information needed to em ploy any KEEL method A full description of the use case files can be found in section 6 KEEL Reference Manual Page 8 of 31 Basic KEEL developement guidelines Developing a new method 1 2 6 Building the executables When the method was fully developed and its relevant configuration files have been created the last step is to add it to the build xml file a ANT script file so the new versions of KEEL could be able to build it inside the KEEL environment The build xml is a critical file so it is not recommended to modify it without authorization of a KEEL project manager
24. mmagt drinks outputs selector data a 92 43 27 317 050 tive 64 59 32 23 lt null gt false 54 lt null gt 16 54 0 0 false 78 34 24 36 0 0 false 55 Nish 177 a sie 62 20 17 9 0 5 true 077 21 WI 11 0 5 true D 22 20 Ya W 5 true 60 25 19 5 0 5 San Sy 24k 1k G2 I UI TO Ss S x x AS oE eF er M v9 0 gp 5 true 5 true KEEL Reference Manual Page 17 of 31 Output files 5 Output files Every method in KEEL must produce at least two output files A train results file marked with the extension tra and a test results file marked with the extension tst Although the method can employ additional output files to show more information about the process performed those additional files must be handled entirely by the method Thus KEEL will only handle the two standards output files Both output files share the same structure They are composite by the same header of the data employed as input of the method and a set of rows one for each instance of the data set describing the expected outputs and the outputs obtained by the application of the method Thus they are structured as follows lt Expected 1 gt lt Expected gt lt Method gt lt Method n gt lt Expected2 gt lt Expected2 gt lt Method gt 1 gt lt Method2 gt 5 1 Example of use As related before the structure of the ou
25. nly available in NOMINAL attributes e convertNominal Value Converts a nominal value to its representa tions as integer an integer between 0 N 1 where N is the number of values defined for the attribute e getDirectionAttribute Returns an integer showing if the attribute is defined as input attribute INPUT output attribute output or undefined DIR NOT DEF e getNewValuesInTest Returns n array with the new values of the attribute observed in test data KEEL Reference Manual Page 31 of 31
26. s storing references about the input and output attributes The order of the attributes stored is the same order than it was found in the input data file Its public methods are e getInputAttributes Returns an array containing all the input At tributes e getOutputAttributes Returns an array containing all the output At tributes e getInputAttribute Returns a single input attribute e getOutputAttribute Returns a single output attribute e getAttribute Returns a single attribute defined neither as input nor as output attribute e getNumInputAttributes Returns the number of input attributes e getNumOutputAttributes Returns the number of output attributes e getNumAttributes Returns the number attributes including input output and undefined ones 7 3 4 Attribute The Attribute class contains the definition an attribute of the dataset Its public methods are e getType Returns a integer value defining the type of the attribute the type is defined as NOMINAL INTEGER or REAL e getName Returns the name of the attribute e getMinAttribute Returns the minimum value of the attribute only available in INTEGER or REAL attributes KEEL Reference Manual Page 30 of 31 API Dataset Description of the classes e getMaxAttribute Returns the minimum value of the attribute only available in INTEGER or REAL attributes e getNominalValuesList Returns an array with all the values defined for the attribute o
27. sections will describe their main characteristics 7 3 1 InstanceSet This class contains a complete set of instances Its public methods are e numInstances Returns the number of instances of the Instance Set e getInstance Returns a concrete instance contained in the Instance Set e getInstances Returns an array with all the instances of the Instance Set 7 3 2 Instance The objects of this class represents instances of the data sets Its pubic methods are e getInputReal Values Returns an array containing all the input values of the instance only the positions with INTEGER or REAL attributes values will produce a value e getInputNominalValues Returns an array containing all the input values of the instance only the positions with NOMINAL attributes values will produce a value KEEL Reference Manual Page 28 of 31 API Dataset Description of the classes e getInputMissing Values Returns a boolean array defining which in put values are missing e getInputRealValue Returns the value of a concrete input attribute only the positions with INTEGER or REAL attributes values will produce a value e getInputNominalValue Returns the value of a concrete input at tribute only the positions with NOMINAL attributes values will pro duce a value e getInputMissing Value Returns a boolean value defining if the input value is missing e getOutputRealValues Returns an array containing all the output values of the
28. some operations have to be performed in order to fully integrate it By following this guidelines a developer can left all the input output operations to be accomplished by KEEL environment focusing its efforts in the construction of the method itself The steps needed to complete the integration of a new method in KEEL are Reading of the configuration file e Development of the method Writing the output files Registering the method in KEEL Making the use case files Building the executables of the method KEEL Reference Manual Page 3 of 31 Basic KEEL developement guidelines Developing a new method 1 2 1 Reading of the configuration file The KEEL methods only accept one parameter The name and path of a configuration file A typical main class of a method can be the following public class Main private static Clas classifier public static void main String args if args length 1 System err println Error else classifier new Clas args 0 classifier execute end method end class The configuration file contains information about the input and output files of the method In addition it contains the values for all the parameters defined A full description of the configuration files can be found in section 3 By interpreting this file the method should be able to acquire the correct values of its parameters including the seed to initialize the
29. tput method is derived from the input data files employed For example if the following file is employed as input data of a method relation banana attribute atl real 3 09 2 81 attribute at2 real 2 39 3 19 attribute class 1 0 1 0 inputs atl at2 outputs class data 1 14 0 114 1 0 1 05 0 72 1 0 0 916 0 397 1 0 1 09 0 437 1 0 0 584 0 0937 1 0 1 83 0 452 1 0 1 25 0 286 1 0 KEEL Reference Manual Page 18 of 31 Output files Example of use A valid output file should be formatted like the following file note the single spacing between columns relation banana attribute atl real 3 09 2 81 attribute at2 real 2 39 3 19 attribute class 1 0 1 0 inputs atl at2 outputs class data 10 i 1 0 1 0 LO O 1 0 1 0 1 0 1 0 1 11 0 1 0 1 0 By employing this structure it is easy to understand how well the task was performed by the method in the example shown above the method failed to predict the output for the instances 2 and 3 and predicted correctly the remaining ones KEEL Reference Manual Use Case files 6 Use Case files The use case files of KEEL provides valuable information to understanding every of the methods which are available to use They are XML files located in the dist help directory Each KEEL use case file is composed by 4 sections lt method gt N
Download Pdf Manuals
Related Search
Knowledge Extraction Evolutionary Learning Reference manual