Skip to content

ahmadali-jamali/Feature-selcetion

Repository files navigation

Feature-selcetion comparision

Feature selection comparison in breath cancer dataset analysis and preprocessing in feature selection and engeniering and modeling with Random Forest. The model has been test with 6 different feature selection method as comparision of accuracy in machine learning model.

The feature selection methods :

  1. Univaraint
  2. PCA
  3. Changing Lasso
  4. recursive_feature_elimination
  5. sequentialfeatureselector_method
  6. model_based_method

As Random Forest has its own way to feature selection by cost function of Entrupy or Gini. For analysis the performances of the feature selections methods Random Forest is compatible benchmark as scoring and comparision of feature selection methods.

This machin learning has been tested for Supervised Binary classification of breath canser data set.