Python package to conduct feature-reweighted representational similarity analysis.

This repository contains the core model we called "Collaborative filtering enhanced Content-based Filtering" published in our UMUAI article "Movie Genome: Alleviating New Item Cold Start in Movie Recommendation"

Source Code for Feature Weighting with Ant Lion optimization in MATLAB

Distributed version of RELIEF-F algorithm for Apache Spark.

A multi-objective optimization project using NSGA2

Applying various feature weighting methods for classifying data using knn

Source Code for Improved Ant Lion optimization to search feature weights. Implemented in MATLAB

Feature-weighted FKNN regression using fuzzy mutual information and Lukasiewicz similarity

Technical working paper on optimizing the weighted nearest neighbors (wNN) regression algorithm by learning feature-specific distance metric weights via batch gradient descent, achieving up to 80.65% MSE reduction over conventional wNN on UCI benchmarks.

This project explores metaheuristic algorithms for feature weight learning in 1NN classification.

Weight whole feature groups by cross-fold contribution to generalization; held-out + magnitude-matched null validated; numpy-only.