Application of Unsupervised Method Based on Evolutionary Neural Network in Evaluation of the Importance of Indexes in Sports Measurement

Unsupervised evaluation about the importance of indexes in sports measurement is a basic problem.Traditional method is often influenced by subjective factors and random fuzzy problems, against this, an unsupervised method based on evolutionary neural network is proposed in this paper to evaluate the importance of indexes.First, with DB rules and fuzzy C means clustering algorithm to determine the optimal number of clusters and clustering results of the sample data set.Secondly, regarding the data set as the neural network training set, optimize the connection weights of neural network with genetic algorithm in order to improve the neural network performance and to get a good connection weights.Finally extract the weights from the trained neural network and convert them to the relative importance of indexes.The result shows that this method can accurately reflects the relative importance of indexes in sports measurement, and also have good maneuverability.