/MLforSW

Primary LanguagePython

MLforSW

Running instructions

  1. Import the libraries
    pip install numpy
    pip install imblearn
    pip install scipy
    pip install pandas
    pip install matplotlib

  2. Data and results stored in /data

  3. Source code stored in /src

  4. Other folder consists of testing code

Classifiers Used

  1. Linear Regression
  2. Naive Bayes Gaussian (same as diagonal quadratic) [theory??]
  3. Generalized linear model
  4. Decision Tree
  5. Extreme Learning Machine (linear kernel)
  6. Extreme Learning Machine (polynomial kernel)
  7. Extreme Learning Machine (RBF kernel)

Final Results

Found in the data directory as CSV files, PDF files and xlsx files.

Rough Work

  1. 7 sets of metrics selection techniques
  1. Table style For each project For Accuracy and AUC For each feature selection, make table and compare CL1 CL2 ... CL7 Without Fold1 Fold2 . . Fold5

With Fold1 Fold2 . . Fold5

  1. ??? Make boxplot, for the AUC results. Descriptive statistics - results tabulation - write insights.

  2. Wilcoxon-test analysis How different are the feature selection technques?

Replace al NaN with 0. <- model does not exist.

Think of how to do the remaining 3 comparisons, you can do it!

  1. Classification

  2. Sampling

PPT Motivation Framework/Flowchart of work Tables