基于不同相似性度量的方法对时间序列进行聚类,比较改进TS-DTW距离与其他距离度量相似性在聚类结果上的效果。结果表明基于改进TS-DTW距离度量的聚类结果比其他方法更有效。利用上海证券交易所50指数成分股进行实证研究,采用改进TS-DTW...基于不同相似性度量的方法对时间序列进行聚类,比较改进TS-DTW距离与其他距离度量相似性在聚类结果上的效果。结果表明基于改进TS-DTW距离度量的聚类结果比其他方法更有效。利用上海证券交易所50指数成分股进行实证研究,采用改进TS-DTW距离进行聚类,聚类结果表明不同类别的股票后续仍具有一定时效性,基于此构建投资组合,得到的时间序列聚类模型有助于降低投资组合的波动风险。Clustering time series based on different similarity metrics compares the effect of improved TS-DTW distance with other distance metrics of similarity in clustering results. The results show that the clustering results based on the improved TS-DTW distance metric are more effective than other methods. An empirical study is carried out using the constituent stocks of Shanghai Stock Exchange 50 Index, and the clustering results using the improved TS-DTW distance indicate that the follow-up of different categories of stocks is still time-sensitive, based on which the investment portfolios are constructed, and the obtained time-series clustering model helps to reduce the volatility risk of the investment portfolios.展开更多
文摘基于不同相似性度量的方法对时间序列进行聚类,比较改进TS-DTW距离与其他距离度量相似性在聚类结果上的效果。结果表明基于改进TS-DTW距离度量的聚类结果比其他方法更有效。利用上海证券交易所50指数成分股进行实证研究,采用改进TS-DTW距离进行聚类,聚类结果表明不同类别的股票后续仍具有一定时效性,基于此构建投资组合,得到的时间序列聚类模型有助于降低投资组合的波动风险。Clustering time series based on different similarity metrics compares the effect of improved TS-DTW distance with other distance metrics of similarity in clustering results. The results show that the clustering results based on the improved TS-DTW distance metric are more effective than other methods. An empirical study is carried out using the constituent stocks of Shanghai Stock Exchange 50 Index, and the clustering results using the improved TS-DTW distance indicate that the follow-up of different categories of stocks is still time-sensitive, based on which the investment portfolios are constructed, and the obtained time-series clustering model helps to reduce the volatility risk of the investment portfolios.
