Advanced Dual-Optimized Neural Network Model with Integrated CSO and OBD for Precise Classification and Prediction of North Indian Light Classical Music Genres

Introduction: categorizing North Indian Light Classical Music genres presents a considerable challenge due to their intricate nature. This research introduces a Dual Optimized Neural Network (DONN) model designed to achieve elevated levels of accuracy and efficiency, thereby enhancing the understanding of these music genres. Creating a network of artificial neural with accurate classification and prediction of given genres is the primary objective. This is achieved through the integration of Cat Swarm Optimization (CSO) for enhanced adaptability and Optimal Brain Damage (OBD) for effective network pruning. Methods: the DONN model employs CSO to investigate the solution space effectively while using OBD to minimize unnecessary network connections, thereby improving both computational efficiency and generalization capabilities. The methodology involves modelling the network using a dataset of North Indian Light Classical Music, optimizing the search process with CSO, and applying OBD for network pruning. Results: the DONN model demonstrated a remarkable 98 % accuracy in classifying eleven distinct genres, outperforming previous methods, and highlighting superior classification accuracy and resilience. Compared to earlier research work and Swarm Optimization like Bat and Ant Colony, and Particle Swarm Algorithm, this model shows higher accuracy and efficiency. The fusion of CSO and OBD significantly enhances performance, improving generalization and reducing computational complexity. Conclusions: overall, the DONN model, optimized with CSO and OBD, significantly advances the classification and prediction of North Indian Light Classical Music genres. This research offers a robust and reliable tool for music classification, contributing to a deeper understanding and appreciation of these genres.