Citation

Dhanke JA, Maurya RK, Navaneethan S, Mavaluru D, Nuhmani S, Mishra N, Venugopal E. Comput. Intell. Neurosci. 2022; 2022: e1359714.

Copyright

(Copyright © 2022, Hindawi Publishing)

DOI

10.1155/2022/1359714

PMID

36210988

PMCID

PMC9546649

Abstract

Artificial intelligence has rapidly grown and has made the scenario that no field can function without it. Like every field, it also plays a vital role in the sports field nowadays. In certain sports, injuries happen very often due to heavy training and sudden speedy actions, especially in athletics and football. Here arises a need to analyze the effect of physical training in sportsperson by collecting data from their daily training. With the help of artificial intelligence, a recurrent neural model is developed to analyze the effect of physical training and treatment concerning sports injury. A Recurrent Neural Network (RNN) can be a subsection of Artificial Neural Networks (ANN) that uses the neural nodes connected in a temporal sequence. The temporal sequence is one of the essential terms in this research, which denotes a data sequence of events in a given timeframe. The recurrent neural model is an intelligent machine learning method that comprises a neural schema replicating humans. This neural schema studies the data it collects from the athletes/players and processes it by analyzing previous injuries. Sports injuries have to be analyzed because, in some cases, it becomes more dangerous to the sportsperson that they may even lose their career due to disability. Sometimes it may cause a massive loss to the club or company that hired the sportsperson for the sport. The prediction process can give the player rest until he recovers, thus becoming the safest approach in sports. Therefore, it is essential to analyze the sportsperson's track data to keep an eye on his health. In this research, RNN model is compared with the existing Support Vector Machine (SVM) in concerning to the effect of physical training and treatment for sports. The results show that the proposed model has achieved 99% accuracy, which is higher than the existing algorithm.

Language: en