BCNRBO: Binary Chaotic Newton Raphson Based Optimizer with Dynamic Potential for Feature Selection Problem

This repository contains the official implementation of the Binary Chaotic Newton Raphson Based Optimizer (BCNRBO) with Dynamic Potential for solving feature selection problems.

The proposed algorithm:

Integrates a Dynamic Potential mechanism based on Hamming distance.

Applies Chaotic mapping strategies to enhance exploration capability.

Adapts the Newton-Raphson search rule for optimization.

Converts continuous solutions into binary representations using a novel transfer function.

Evaluates performance using K-fold cross-validation and classification error.

The algorithm is designed to select compact and informative feature subsets while maintaining high classification accuracy.
The related research article is currently under review at Scientific Reports.
Bibliographic information will be updated upon publication.