1. Introduction
Feature Selection Techniques help to select features which contribute most to your prediction variable or output. Or they help to remove irrelevant or partially relevant features (noise).
2. Benefits
- Reduces Over-fitting (less noise)
3.Techniques:
3.1 Evaluate the importance of feature
Select features according to the k highest scores.
Feature importances with forests of trees.
Refer: Feature selection
3.2 Find the correlation between features
Using Correlation Matrix.
Feature Selection Techniques help to select features which contribute most to your prediction variable or output. Or they help to remove irrelevant or partially relevant features (noise).
2. Benefits
- Reduces Over-fitting (less noise)
- Improves Accuracy (less noise)
- Reduces Training Time (less features)3.Techniques:
3.1 Evaluate the importance of feature
Select features according to the k highest scores.
Feature importances with forests of trees.
Refer: Feature selection
3.2 Find the correlation between features
Using Correlation Matrix.
0 Comments