摘要
股票价格在一段时间内的波动趋势对研究预测这支股票有着非常重要的作用。将K-means聚类利用层次聚类算法思想进行改进,对股价波动趋势进行聚类,并在一定范围内选择出最优聚类数。此方法在此问题上聚类效果优于原始k-means聚类算法,为未来对股价波动趋势的预测研究提供帮助。
The fluctuation of stock price in a period of time is very significant to study and forecast the stock.In this paper,we use the Hierarchical clustering thought to improve K-means clustering,and to custer stock price fluctuation trend.Then select the optimal cluster number within a certain range.The clustering effect is better than that of the original K-means clustering algorithm,which will help to forecast the trend of stock price in the future.
出处
《科技和产业》
2016年第1期144-148,共5页
Science Technology and Industry
关键词
股价波动趋势
K均值聚类
层次聚类
stock price fluctuation trend
K-means
hierarchical clustering
