{"title":"Investigation on Performance of Change Point Algorithm in Time Series Dynamical Regimes and Effect of Data Characteristics","authors":"Farhad Asadi, Mohammad Javad Mollakazemi","volume":94,"journal":"International Journal of Bioengineering and Life Sciences","pagesStart":1787,"pagesEnd":1794,"ISSN":"1307-6892","URL":"https:\/\/publications.waset.org\/pdf\/10000464","abstract":"<p>In this paper, Bayesian online inference in models of<br \/>\r\ndata series are constructed by change-points algorithm, which<br \/>\r\nseparated the observed time series into independent series and study<br \/>\r\nthe change and variation of the regime of the data with related<br \/>\r\nstatistical characteristics. variation of statistical characteristics of time<br \/>\r\nseries data often represent separated phenomena in the some<br \/>\r\ndynamical system, like a change in state of brain dynamical reflected<br \/>\r\nin EEG signal data measurement or a change in important regime of<br \/>\r\ndata in many dynamical system. In this paper, prediction algorithm<br \/>\r\nfor studying change point location in some time series data is<br \/>\r\nsimulated. It is verified that pattern of proposed distribution of data<br \/>\r\nhas important factor on simpler and smother fluctuation of hazard<br \/>\r\nrate parameter and also for better identification of change point<br \/>\r\nlocations. Finally, the conditions of how the time series distribution<br \/>\r\neffect on factors in this approach are explained and validated with<br \/>\r\ndifferent time series databases for some dynamical system.<\/p>\r\n","references":"[1] Yu AJ, Dayan P, \u201cUncertainty, neuromodulation, and attention,\r\nNational Center for Biotechnology Information\u201d, 19;46(4):526-8,2005\r\nMay.\r\n[2] Timothy E J Behrens, Mark W Woolrich,\u201d Learning the value of\r\ninformation in an uncertain world\u201d. Nature Neuroscience 10, 1214 -\r\n1221, 2007. [3] Xiang Xuan, \u201cModeling changing dependency structure in multivariate\r\ntime series\u201d, Proceedings of the 24th international conference on\r\nMachine learning, Pages 1055-1062, 2004.\r\n[4] Kiyohito Iigaya,\u201d Dynamical Regimes in Neural Network Models of\r\nMatching Behavior\u201d neural computation, Vol. 25, No. 12, Pages 3093-\r\n3112 ,December 2013.\r\n[5] M Steyvers,\u201d prediction and change detection\u201d: Springer-Verlag, 2006.\r\n[6] David J.C. MacKay, \u201cEnsemble Learning for Hidden Markov Models\u201d,\r\n1997.\r\n[7] Adams, R. P., & MacKay, D. J., \u201cBayesian online changepoint\r\ndetection.\u201d (Tech. Rep.). Cambridge: Cambridge University, 2007.\r\n[8] Fearnhead, P., & Liu, Z \u201cOn-line inference for multiple changepoint\r\nproblems\u201d. Journal of the Royal Statistical Society: Series B (Statistical\r\nMethodology), 69(4), 589\u2013605.\r\n[9] Xuan, X., & Murphy, K., Rocky ,\u201d Modeling changing dependency\r\nstructure in multivariate time series. In Proceedings of the 24th\r\nInternational Conference on Machine Learning. San Francisco: Morgan\r\nKaufmann, 2007.\r\n[10] Wilson, R. C, \u201cParallel Hopfield networks.\u201d Neural Computation, 21(3),\r\n831-850, 2009.\r\n[11] Wilson, R. C., Nassar, M. R., & Gold, J. I, Bayesian online learning of\r\nthe hazard rate in change-point problems. Neural Computation, 22(9),\r\n2452-2476, 2010.\r\n[12] Farhad Asadi, Mohammad javad Mollakazemi et all, \u201cThe influence of\r\nparameters of modeling and data distribution for optimal condition on\r\nlocally weighted projection regression method\u201d.. Accepted and oral\r\npresentation in ICMSE 2014: XII International Conference on\r\nMathematics and Statistical Engineering, October, 27-28, 2014, Istanbul,\r\nturkey.\r\n[13] Mohammad javad Mollakazemi, Farhad Asadi., \u201cReal-time adaptive\r\nobstacle avoidance in dynamic environments with different D-S\u201d.\r\nAccepted and oral presentation in ICARM 2014: XII International\r\nConference on Automation, Robotics and Mechatronics, October, 27-28,\r\n2014, Istanbul, turkey.","publisher":"World Academy of Science, Engineering and Technology","index":"Open Science Index 94, 2014"}