The application of artificial intelligence technology in the tactical training of football players

Abstract

Tactical training for football players involves ground-level goal concentration, player passing, etc., as features. Tactical training is mandatory to improve the skills of a player in terms of performance and reasoning ability for goals. This work describes a performance-focused strategic training module (PFSTM) that uses a concatenated learning network. The performance-focused player features are identified via different stats from those used in the previous games. The lagging features toward in-ground performance are identified via concatenated outcomes to provide specific tactical training. In such training sessions, the tactical training features are determined via recurrent learning based on the maximum achievable training outcomes. The concatenation between the features overlaps under multiple strategy evaluation sessions to maximize player performance. In this case, the concatenation of performance features is segmented and released after their improvements. Thus, the proposed module is designed to fit tactical training and in-ground application of football players irrespective of the sessions and strategies. From previous training, this module identifies various performance features, such as ball passing accuracy, awareness of handling the ball, and strategies made by the players. Players with better fitness levels will quickly improve skills such as ball passing and reaching target accuracy. The results demonstrate an accuracy of 95 % and an RMSE of 0.15, indicating strong predictive capabilities.