Introduce the principle and the theory below the Correlation based Feature Selection method.
CFS: Correlation-based Feature Selection is composed with three parts:
- Feature Evaluation
The heart of CFS algorithm is a heuristic for evaluation the worth or merit of a subset of features;
- Feature Correlations
Information gain is used to calculate the correlation between different features and class;
- Searching the Feature Subset Space
CFS starts from the empty set of features and uses a forward best first search with a stopping criterion of five consecutive fully expanded non-improving subsets.