This project provides a comprehensive toolset for feature selection using LightGBM, a gradient-boosting framework that uses tree-based learning algorithms. The primary goal is to improve model performance by selecting the most relevant features and discarding the redundant ones.
Description: This directory contains custom data loaders for classification and time series data.
classification_dataloader.py: For loading classification datasets.time_series_dataloader.py: For loading time series datasets.m4_feature_extractor.py: Extracts features from the M4 dataset.
Description: This directory houses implementations of different machine-learning models and pipelines.
Feature_Selector_LightGBM.py: Feature selection using LightGBM.Feature_Selector_MLP.py: Feature selection using MLP.LightGBM_Pipeline.py: A pipeline for LightGBM.MLP_Pipeline.py: A pipeline for MLP.XGBoost_Pipeline.py: A pipeline for XGBoost.
Description: (The content didn't provide specific details about this directory. You might want to describe what kind of results or output files are stored here.)
Description: This directory contains scripts to run the feature selector models.
LightGBM_Feature_Selector_Runner.py: Runner for the LightGBM feature selector.MLP_Feature_Selector_Runner.py: Runner for the MLP feature selector.
-
Clone the repository: git clone https://github.com/YigitTurali/AFS_BM-Algorithm.git
-
Navigate to the repository directory: cd AFS_BM-Algorithm
-
Install the required packages: pip install -r requirements.txt
- Load your dataset using the appropriate data loader from the
DataLoadersdirectory. - Choose the model or pipeline you want to use from the
Modelsdirectory. - Run the corresponding runner script from the root directory to start the feature selection process.
Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.
This project is licensed under the MIT License.