Research on learning and evaluation model of traditional sports skills of ethnic minorities using optimized spiking deep residual network

Abstract

The athletic performance of competitors in sporting events is influenced by a range of parameters. It is challenging for coaches to develop specialized, scientific training programs for athletes’ problems because the standard sports performance prediction approach makes it impossible to obtain precise findings and the related data analysis promptly. A sports performance prediction using the Spiking Deep Residual Network based Coati Optimization Algorithm (SPP-SDRN-COA) is proposed in this manuscript. Initially, the input data was collected from the historical data of ethnic minority athletics. Afterward, data was fed to pre-processing. In pre-processing, the Domain Transform Filtering (DTF) method removed noise by using DTF. Next, the pre-processed data was given to the Spiking Deep Residual Network (SDRN) which classified the sports performance of an athlete’s strengths and weakness. In general, SDRN does not show any optimization adaption methods to determine the optimum parameter to offer an accurate prediction. The Coati Optimization Algorithm (COA) is proposed in this manuscript to optimize the SDRN classifier that detects the strengths and weaknesses of the sports performance in an accurate manner. The proposed SPP – SDRN – COA is implemented using the MATLAB platform. By using the proposed SPP – SDRN – COA based sports performance prediction, the method provides higher accuracy and better prediction. Additionally, coaches can analyze athletic performance and gain a more thorough understanding of athletes’ strength levels than using the conventional performance prediction method.