Feature Engineering Technique: Python Selection and Transformation

NhanPhamThanh-IT released this repository in August 2025, providing a collection of datasets and Python-based workflows for feature engineering and selection. It includes Jupyter notebooks covering EDA, data cleaning, and five distinct feature selection methods like filter, wrapper, and hybrid approaches.

Use Cases

• Applying discretization and encoding techniques to categorical features within the provided datasets
• Executing filter, wrapper, and embedded feature selection methods to identify high-impact variables
• Handling missing data and outliers using the provided Python utility functions

Strengths

• Covers 5 distinct feature selection categories: filter, wrapper, embedded, shuffling, and hybrid
• Includes reusable Python utility scripts for automated feature transformation
• MIT licensed for open commercial and academic use

Limitations

• Specific row counts and column definitions for the included datasets are not documented in the metadata
• Primary focus is on educational implementation rather than a single large-scale production dataset

Provenance

Source

NhanPhamThanh-IT via GitHub

Freshness

Last updated August 2025.