Title :
A Fast Two-Stage Dynamic Programming Algorithm for Change-Points Model with Application in Speech Signal
Author :
Chang, Kuo-Ching ; Chiang, Chui-Liang ; Lee, Chung-Bow
Author_Institution :
Dept. of Appl. Math., Nat. Chung Hsing Univ., Taichung, Taiwan
Abstract :
In multiple change-points model, the dynamic programming (DP) algorithm can be use to obtain the maximum likelihood estimation for a sequence of data from multivariate normal distribution. Since the algorithm has a quadratic complexity in data size n, it is computationally burdensome if the data size n is large. In this paper we present a fast two-stage dynamic programming (TSDP) through the window method. In TSDP algorithm, the first stage is to use the window method based on the log-likelihood ratio measure to find a subset of candidate change points. The second stage is to apply DP algorithm on the chosen subset to detect the position of change points. The proposed algorithm of change-points will be used for the boundary detection of speech signal by finding the abrupt spectral difference change of adjacent frames. Some simulated data sets and the speech data are investigated for DP and TSDP algorithms. In comparison of CPU times, the TSDP algorithm can be up to 34.96 and 74.02 times faster than the DP algorithm for the simulated data and the speech data respectively. The results show that our algorithm works very well. It substantially reduces the computation load for large data size n.
Keywords :
dynamic programming; maximum likelihood estimation; normal distribution; quadratic programming; speech processing; CPU times; TSDP algorithm; boundary detection; computation load; data sequence; log likelihood ratio measurement; maximum likelihood estimation; multiple change point model; multivariate normal distribution; quadratic complexity; speech data; speech signal; two stage dynamic programming algorithm; window method; Complexity theory; Computational efficiency; Data models; Dynamic programming; Heuristic algorithms; Maximum likelihood estimation; Speech; Change points; Dynamic programming; Multivariate normal density; Speech signal;
Conference_Titel :
Genetic and Evolutionary Computing (ICGEC), 2010 Fourth International Conference on
Conference_Location :
Shenzhen
Print_ISBN :
978-1-4244-8891-9
Electronic_ISBN :
978-0-7695-4281-2
DOI :
10.1109/ICGEC.2010.97