Martial Arts Competitive Decision-Making Algorithm Based on Improved BP Neural Network

Area of Science:

• Sports Science
• Artificial Intelligence
• Algorithm Development

Background:

• Modern Chinese martial arts competitions heavily rely on strategy, necessitating advanced decision-making algorithms.
• Effective strategy formulation requires scientific analysis of athletic data and accurate prediction capabilities.
• Traditional algorithms face challenges in adapting learning rates and controlling weight adjustments.

Purpose of the Study:

• To develop a novel algorithm for constructing Chinese martial arts competition decision-making strategies.
• To improve the prediction accuracy of martial arts competition routines.
• To enhance the scientific formulation of strategies in sports competitions.

Main Methods:

• Implementation of a novel additional momentum-elastic gradient descent algorithm.
• Adaptation of the Backpropagation (BP) neural network for learning rate adjustment.
• Improvement of traditional BP neural network parameters like learning step length, weight direction, and size.

Main Results:

• The proposed algorithm demonstrates improvements in network scale and running time.
• The algorithm successfully predicts martial arts competition routines.
• Scientific strategies for martial arts competitions can be formulated using this method.

Conclusions:

• The developed algorithm offers a more effective approach to strategy formulation in Chinese martial arts competitions.
• Neural network technology, particularly the enhanced BP network, is valuable for sports data analysis and prediction.
• This research contributes to the scientific advancement of martial arts competition strategy and decision-making.