Abstract
Regression is one of the important problems in statistical learning theory. This paper proves the global convergence of the piecewise regression algorithm based on deterministic annealing and continuity of global minimum of free energy w.r.t temperature, and derives a new simplified formula to compute the initial critical temperature. A new enhanced piecewise regression algorithm by using “migration of prototypes” is proposed to eliminate “empty cell” in the annealing process. Numerical experiments on several benchmark datasets show that the new algorithm can remove redundancy and improve generalization of the piecewise regression model.
Similar content being viewed by others
Reference
Rose K. Deterministic annealing for clustering, compression, classification, regression, and related optimization problems. Proc IEEE, 1998, 86(11): 2210–2239
Rose K, Gurewitz E, Fox G. A deterministic annealing approach to clustering. Patt Recog Lett, 1990, 11: 589–594
Persson P, Nordebo S, Claesson I. Design of discrete coefficient FIR filters by a fast entropy-directed deterministic annealing algorithm. IEEE Trans Sig Proc, 2005, 53(3): 1006–1014
Koulgi P, Regunathan S L, Rose K. Multiple description quantization by deterministic annealing. IEEE Trans Inf Theo, 2003, 49(8): 2067–2075
Rao A V. A deterministic annealing approach for parsimonious design of piecewise regression models. IEEE Trans Patt Anal Mach Intell, 1999, 21(2): 159–172
Hastie T, Tibshirani T, Friedman J, The Elements of Statistical Learning: Data Mining, Inference and Prediction. New York: Springer-Verlag, 2001
Briedman L. Classification and Regression Tree. Belmont, CA: Wadsworth Inc., 1984
Rao A V. Mixture of experts regression modeling by deterministic annealing. IEEE Trans Sig Proc, 1997, 45(11): 2811–2820
Author information
Authors and Affiliations
Corresponding author
Additional information
Supported by the National Natural Science Foundation of China (Grant Nos. 60675013 and 4022500) and the National Basic Research Program of China (973 Program) (Grant No. 2007CB311002)
Rights and permissions
About this article
Cite this article
Zhang, J., Yang, Y., Chen, X. et al. Enhanced piecewise regression based on deterministic annealing. Sci. China Ser. F-Inf. Sci. 51, 1025–1038 (2008). https://doi.org/10.1007/s11432-008-0079-0
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11432-008-0079-0