Clustering framework based on multi-scale analysis of intraday financial time series
Yong Shia b c d, Bo Lib c d✉, Guangle Due f, Wei Daib c d
- aCollege of Electrical & Information Engineering, Southwest Minzu University, Chengdu, 610041, China
- bKey Laboratory of Big Data Mining and Knowledge Management, Chinese Academy of Sciences, Beijing, 100190, China
- cResearch Center on Fictitious Economy and Data Science, Chinese Academy of Sciences, Beijing, 100190, China
- dSchool of Economics and Management, University of Chinese Academy of Sciences, Beijing, 100190, China
- eSchool of Physical Sciences, University of Chinese Academy of Sciences, Beijing 100049, China
- fWenzhou Institute, University of Chinese Academy of Sciences, Wenzhou, Zhejiang 325001, China
- ✉Corresponding author
Papers open inside SciRah from the original publisher or repository; SciRah never hosts them.
Abstract
Abstract The analysis of intraday financial time series is the basis of constructing intraday trading strategies which are usually less risky than overnight trading strategies. Correlations existed in intraday financial series may imply there are some potential patterns of price movements. In this work, we propose a clustering framework based on multi-scale analysis of intraday financial time series to seek these potential patterns. The clustering framework include a new method based on multi-scale analysis of time series to measure the similarity between intraday financial time series, and quantitative indexes constructed to evaluate the clustering effect of intraday financial time series. We use different types of clustering algorithms to verify our clustering framework on the China Securities Index 300 (CSI 300), the Standard & Poor’s 500 index (S&P 500) and the Nikkei 225 index (N225), and find that our proposed framework can clearly distinguish different classes of intraday financial time series.
TL;DRA clustering framework based on multi-scale analysis of intraday financial time series to seek potential patterns of price movements is proposed and it is found that the proposed framework can clearly distinguish different classes ofintradayfinancial time series. (Semantic Scholar)