Optimal Model of Horse Racing Competition Decision Management Based on Association Rules and Neural Network

Abstract

With the vigorous development of horse racing, people’s attention to horse racing has increased significantly. Some experts and scholars have conducted research on the decision-making management and predictive analysis methods of horse racing. Today, with the rapid development of information technology, the amount of data and data dimensions of horse racing competitions continue to explode. The increase in data scale and feature dimensions provides new challenges for competition management and competition prediction research. At present, traditional prediction algorithms can no longer meet the needs of horse racing situation prediction, but research has found that association rules and neural network algorithms provide a good solution to the classification and prediction problem. Based on the advantages of association rules and neural networks in analyzing data, according to the requirements of horse racing decision management, this paper adopts the B/S structure to realize the construction of the horse racing decision management optimization model from the three aspects of hierarchical structure, functional structure, and forecasting process. Combined with the horse racing decision management optimization model, based on a large number of experimental training data, the final conclusion is drawn: first, the factors that affect the horse racing performance are from large to small. The order of arrangement is: race schedule > age > gender > weight > rating > horse top three rate > jockey > weight load > harness > ranking > field nature > field > trainer; the second is the prediction and actual results of the neural network algorithm. The closest one, which is slightly higher than 90%, has the highest prediction accuracy; third, the average value of the horse racing performance prediction of this system during the review is only 2.01 s, and the misappraisal rate is 0.12%, indicating that the application value of this system is significant; fourth, in the average time spent in the two seasons, the average time spent in the second season was reduced compared with the average time spent in the first season, with a maximum reduction of 0.984 s, indicating a slight improvement in the performance of the 2020 season. Using this system to predict horse racing, results can improve the optimization of horse racing decision management to a certain extent.