File	Date	Author	Commit
lib	2016-04-04	mohammad	[df8682] Release Candidate 1
nbproject	2016-04-04	mohammad	[df8682] Release Candidate 1
src	2016-04-04	mohammad	[df8682] Release Candidate 1
.gitignore	2016-04-04	mohammad	[df8682] Release Candidate 1
README.TXT	2016-04-04	mohammad	[df8682] Release Candidate 1
build.xml	2016-04-04	mohammad	[df8682] Release Candidate 1
manifest.mf	2016-04-04	mohammad	[df8682] Release Candidate 1

Read Me

========================
GA-EoC for PloS One
========================

GA-EoC: Genetic Algorithm-based Search Method for Heterogeneous Ensembles
Version-1.1: Uses Prebuilt CV Datasets and CV Models for finding Best Ensembles using the Training Dataset.

Required Libraries:
===================
weka: we use Waikato Environment for Knowledge Analysis (WEKA) version 3.7.10. The jar file for required version has been included in the "lib" directory. Alternatively, you can download it from publisher's webaite at http://www.cs.waikato.ac.nz/ml/weka/index.html
libsvm for weka: For using the LibSVM classifier with weka, we have used  Wrapper class for the libsvm library by Chih-Chung Chang and Chih-Jen Lin (http://www.csie.ntu.edu.tw/~cjlin/libsvm/). The original wrapper, named WLSVM, was developed by Yasser EL-Manzalawy (Yasser EL-Manzalawy (2005). WLSVM. URL http://www.cs.iastate.edu/~yasser/wlsvm/). We have included the required wrappers for using the LibSVM inside the "lib" directory.
java: we used JDK 1.6 for the development of the souce code. If you run the program in a multi-core computer, it will leave 2 processor cores free and use rest cores for this program.

How to Run:
============
1). Unzip the GA-EoC.zip file into a folder.
2). To execute or run the program from the command line, go to the "lib" folder and type the following command depending on your intended usage:

Usage 1: Generate CV Models for use in GA-EoC.
java -jar GA-EoC.jar <-options optValue>
-t = input trainFileName ../path/name.
-b = enable flag Build CV Models (no values required).
-d = output CV Data path.
-m = output Prebuilt CV Models path.
-f = CV folds (optional, default 10-fold cv).

Usage 2: Generate Full Models for use in GA-EoC.
java -jar GA-EoC.jar <-options optValue>
-t = input trainFileName ../path/name.
-M = output Path for Full Models.

Usage 3: Use Prebuilt Models to find Best Ensemble Combination.
java -jar GA-EoC.jar <-options optValue>
-t = input trainFileName ../path/name.
-d = input CV Data path.
-m = input Prebuilt Model path.
-r = Repeat the Process (optional, default 50 repeatations).
-l = output Run Logs.
-f = CV folds (optional, default 10-fold cv).

Usage 4: Evaluate the performance of an Ensemble Combination.
java -jar GA-EoC.jar <-options optValue>
-t = input trainFileName ../path/name.
-T = input TestFileName ../path/name.
-M = input Path for Full Models.
-e = input Ensemble Combination.