Title :
A Phylogenetic Analysis for Stock Market Indices Using Time-Series String Alignments
Author :
Kim, Hyong-Jun ; Ryu, Chang-Keon ; Ryu, Dong-Sung ; Cho, Hwan-Gue
Author_Institution :
Dept. of Comput. Sci. & Eng., Pusan Nat. Univ., Busan
Abstract :
Predicting stock is very difficult and complex since there are lots of variables we should consider. Most of stock system usually deals with the stock indices and their properties. But recent stock system can not provide clearly the classification according to similar stock properties. So we need smart clustering algorithm for stock market that can derive their hidden interactions from stock indices. In this paper, we propose phylogenetic analysis for stock market using the string alignments based on time-series to cluster stock indices. We also analyze KOSPI data from 2007/01/03 to 2008/02/28 using string alignments, our proposed algorithms can show the meaningful clusters which can not displayed by the recent other system.
Keywords :
stock markets; time series; KOSPI data; phylogenetic analysis; smart clustering algorithm; stock market indices; time-series string alignments; Algorithm design and analysis; Clustering algorithms; Computer science; DNA; Information analysis; Information technology; Phylogeny; Sequences; Stock markets; Time series analysis; Global Alignment; Stock clustering;
Conference_Titel :
Convergence and Hybrid Information Technology, 2008. ICCIT '08. Third International Conference on
Conference_Location :
Busan
Print_ISBN :
978-0-7695-3407-7
DOI :
10.1109/ICCIT.2008.227