scikit-rebate
Feature selector
An implementation of Relief-based feature selection algorithms for Machine Learning.
A scikit-learn-compatible Python implementation of ReBATE, a suite of Relief-based feature selection algorithms for Machine Learning.
413 stars
25 watching
73 forks
Language: Python
last commit: over 3 years agoLinked from 2 awesome lists
data-sciencefeature-selectionpython
Related projects:
| Repository | Description | Stars |
|---|---|---|
| A collection of algorithms and tools for selecting the most informative features from datasets in machine learning applications. | 1,513 | |
| An implementation of a feature selection method to identify all relevant features for prediction | 1,529 | |
| A software toolbox for feature selection algorithms | 71 | |
| An algorithm for fast feature selection using XGBoost and other tree-based classifiers | 219 | |
| Automates feature engineering by using genetic programming to select the most useful features for machine learning models. | 51 | |
| A genetic feature selection tool for machine learning models | 323 | |
| A Python library that performs feature selection using various nature-inspired optimization algorithms. | 245 | |
| An algorithm for feature construction that models higher-order interactions and captures complex patterns in data sets | 126 | |
| Automated hyperparameter tuning and feature selection using evolutionary algorithms. | 316 | |
| A Python machine learning module that generates logical rules to predict class labels with high precision | 624 | |
| Automated machine learning protocols for cheminformatics using Python | 39 | |
| Provides a solid base for redesigning the native select element using SUIT CSS components | 12 | |
| A Python implementation of a feature point transform designed to be invariant to certain transformations. | 487 | |
| An R package for computing Shapley values to analyze feature contributions in machine learning models. | 25 | |
| A Vue.js component for displaying and selecting multiple items in a grid-like interface. | 38 |