A Novel Hybrid Algorithm for Discovering Motifs from Financial Time Series

Time series motifs are pairs of individual subsequences, which are very similar to each other within the time series. In this paper, we propose an efficient and novel hybrid algorithm by taking best out of two popular algorithms namely MK algorithm and EP-C algorithm that exist in literature. We demonstrate the efficiency of our approach in terms of time elapsed and distance obtained between the subsequences through experiments conducted on financial time series datasets viz., foreign exchange rates, Gold price and Crude oil price in terms of US dollars.

[1]  Duong Tuan Anh,et al.  Time Series Prediction Using Motif Information , 2012, MIWAI.

[2]  Eamonn J. Keogh,et al.  Exact Discovery of Time Series Motifs , 2009, SDM.

[3]  Chun-Ping Li,et al.  Stock temporal prediction based on time series motifs , 2009, 2009 International Conference on Machine Learning and Cybernetics.

[4]  Uwe Aickelin,et al.  The motif tracking algorithm , 2008, Int. J. Autom. Comput..

[5]  Jessica Lin,et al.  Finding Motifs in Time Series , 2002, KDD 2002.

[6]  Bernhard Sick,et al.  On-line motif detection in time series with SwiftMotif , 2009, Pattern Recognit..

[7]  Stephen Shaoyi Liao,et al.  Discovering original motifs with different lengths from time series , 2008, Knowl. Based Syst..

[8]  Jeremy Buhler,et al.  Finding motifs using random projections , 2001, RECOMB.

[9]  Vit Niennattrakul,et al.  Discovery of variable length time series motif , 2011, The 8th Electrical Engineering/ Electronics, Computer, Telecommunications and Information Technology (ECTI) Association of Thailand - Conference 2011.

[10]  Eamonn J. Keogh,et al.  Probabilistic discovery of time series motifs , 2003, KDD '03.