tremila/fstoneadp

FS

java: java codes

• ArffData: Training and testing datasets
• lib: The jar packages that the algorithm needs to call (weka.jar, SVMAttributeEval.jar, chiSquaredAttributeEval.jar, etc)
• src: Java code files
  • cn.lwp.whut.algorithm
    • DataProcess.java: Process the datasets
    • SelectionFeatures.java: Select features with filter-based feature ranking methods, filter-based feature subset selection methods
  • cn.lwp.whut.feature
    • FilterFeatureRanking.java: Filter-based feature ranking methods
    • FilterSubsetSelection.java: Filter-based feature subset selection methods
    • SearchMethod.java: Search strategies for Filter-based feature subset selection methods
  • cn.lwp.whut.main
    • Mian.java: The main method to run the whole java codes

python: python codes

• CrossVersionData: Training and testing datasets pairs
• lib: The packages that the algorithm needs to call (sklearn, numpy, pandas, etc)
  • packages.zip: The compressed packages (sklearn, numpy, pandas, etc). It is recommended to download this 'packages.zip' separately here.
• src: Python code files
  • configuration.py: Store the file paths
  • classifiers.py: The classification models we studied
  • PerformanceMeasure.py: Calculate the performance measures
  • Processing.py: Process the datasets
  • wrapper_subset_selection.py: Define the wrapper-based feature subset selection methods
  • run.py: the main method to run the whole python codes