Research On K-means Clustering Algorithm And Its Application In Stock Investment

Clustering algorithm is one of the data mining algorithms.The most common clustering algorithm is K-means clustering algorithm,but the algorithm is a local optimal algorithm.Only the initial clustering center is properly selected,the algorithm can achieve global optimality.There are many improved methods at home and abroad for the selection of initial clustering centers in K-means clustering algorithm.This thesis has carried on the research and discussion on the basis of previous studies.The research work mainly includes:(1)A new K-means clustering algorithm for selecting the initial clustering center is proposed.The k-dist value is combined with the nearest neighbor method to determine the initial clustering center.The experimental results show that the clustering effect of the new algorithm is better than the traditional K-means clustering algorithm,which is reflected in the larger Sil index and smaller sum of squares of data distances within a class.(2)Applying the new algorithm to stock investment,the results show that the improved new algorithm can classify stocks more efficiently.