Intelligent Prediction and Training Optimization of Sports using Enhanced Whale Optimized Artificial Neural Network

Enhancing the efficiency of the sports training should be provided since many of the athletic training methods used today are ineffective at considering the unique qualities of individual athletes. The proposed research aimed at developing intelligent prediction and optimization of basketball training using data from Federation Internationale de Basketball (FIBA) website and pre-processed using min max normalization. The Capsule Network (CapsNet) of Convolutional Neural Network (CNN) is used to extract features and the classified using Attention mechanism combined Convolutional Random Forest (AttConv-RF). CapsNet makes it possible to improve the model's capabilities due to the updated loss function. The ConvRF unit recognizes spatial and temporal relationships of the training data, decreasing overfitting issues primarily while training data are limited. The performance prediction is carried out using Enhanced Whale Optimization Algorithm based Artificial Neural Network (EWOA-ANN). The results showed that proposed EWOA-ANN approach achieved better performance prediction at accuracy with 99.56%, precision of 97.42%, and recall 96.36% when compared with the existing methods Back Propagation Neural Network (BPNN) and eXtreme Gradient Boost (XGB).