https://github.com/bullet-ant/feature-selection
Implemented 3 ranking techniques: Fisher Discriminant Ratio (FDR), Signal-Noise Ratio (SNR), Corelation Coefficient (CC). Made different combinations of the above 3 to eliminate low ranked features and identify the prominent features and compare the performances of models trained with every combination’s selected features.
https://github.com/bullet-ant/feature-selection
Last synced: about 2 months ago
JSON representation
Implemented 3 ranking techniques: Fisher Discriminant Ratio (FDR), Signal-Noise Ratio (SNR), Corelation Coefficient (CC). Made different combinations of the above 3 to eliminate low ranked features and identify the prominent features and compare the performances of models trained with every combination’s selected features.
- Host: GitHub
- URL: https://github.com/bullet-ant/feature-selection
- Owner: bullet-ant
- License: mit
- Created: 2020-09-13T14:31:16.000Z (over 4 years ago)
- Default Branch: master
- Last Pushed: 2020-09-13T16:33:36.000Z (over 4 years ago)
- Last Synced: 2023-04-29T17:39:15.411Z (about 2 years ago)
- Language: Jupyter Notebook
- Size: 1.69 MB
- Stars: 3
- Watchers: 1
- Forks: 0
- Open Issues: 0