Abstract
In the thesis, we use the time sequence of basketball competition as the statistical data to predict the technical statistical indicators for the basketball team based on BP neural network algorithm. Our study aims at discussing the availability and applications of the BP neural network in the score prediction. Introduction Prediction of competition scores is the prediction of the future sports scores based on the existing sport scores. Scientific prediction of competitions can not only provide athletes and coaches with certain targets for training and competitions, but also track and judge the tendency and regulations of sport items. Therefore, the prediction of competition scores has increasingly become crucial in scientific sport managements and competitions. Also, it has been concerned by the leaders, managers and researchers during current days. With the development of dynamic system theories, artificial neural networks (ANNs) and expert system, sports predictors start to use better approaches that suits for human mind and the dynamic changes of predictive environments. ANNs can learn existing data automatically without too much complex processes and get close to the best function that can classify the samples [1]. Artificial expert system can gather the knowledge and experiences of human and give quantitative results, which makes the problems be more close to actual issues. Therefore, ANNs and artificial intelligence have successfully extend the potential development of predictive systems [2]. Here, we use the time sequence of basketball competition as the statistical data to predict the technical statistical indicators for the basketball team based on BP neural network algorithm. Our study aims at discussing the availability and applications of the BP neural network in the score prediction. Algorithm and Model Development of BP Neural Network BP Neural Network BP neural network is the feed-forward neural network with the transmission of errors. It is usually made up of the input, output and several hidden layers. Each layer consists of a series of nodes, which represents the neurons [3]. The upper layer connects with the utter layer via the weights. Different layers interconnect with each other and there is no connection among neurons that in the same layer. A typical BP network has three layers, including one hidden layer, which is shown in Figure 1:
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have