Advancing multi-class classification: innovations, challenges, and ethical perspectives in machine learning
This review examines recent advances and persistent challenges in multi-class classification within machine learning (ML) and deep learning (DL) and identifies key future directions, including semi-supervised and transfer learning, few-shot and federated multi-class systems, robustness under label noise, energy-efficie...