MCMC GAN for Oversampling to Overcome Class Imbalance

Overview

This repository contains Python code implementing a Markov Chain Monte Carlo (MCMC) Generative Adversarial Network (GAN) for oversampling to address class imbalance in datasets. The approach combines MCMC sampling techniques with the GAN architecture to generate synthetic samples for minority classes, thereby mitigating class imbalance issues commonly encountered in machine learning tasks.

Dependencies Python (>=3.6) NumPy TensorFlow (>=2.0)