Firstly, the similarity measure of stock turnover time series is introduced. By using the fuzzy clustering algo rithm, the stocks which have good correlation and close turnover magnitude are clustered. Then, an empirical analysis , taking the non-ferrous metal stocks as a case study, is conducted , and the algorithm with MATLAB is implemented. Finally, the clustering results are worked out, and the effectiveness of the clustering results is verified through market experience and intution.%引入了股票成交额时间序列的相似性度量，通过模糊聚类方法，将成交额时间序列相关性好、数量级接近的股票聚类，并以有色金属类股票为例进行了实证分析，通过MATLAB编程实现算法，计算出聚类结果，并从市场经验和直观角度对结果进行了分析，说明了聚类结果的有效性．
展开▼
