×

A comparison of estimators for regression models with change points. (English) Zbl 1255.62194

Summary: We consider two problems concerning locating change points in a linear regression model. One involves jump discontinuities (change-point) in a regression model and the other involves regression lines connected at unknown points. We compare four methods for estimating single or multiple change points in a regression model, when both the error variance and regression coefficients change simultaneously at unknown point(s): Bayesian, Julious, grid search, and the segmented methods. The proposed methods are evaluated via a simulation study and compared via some standard measures of estimation bias and precision. Finally, the methods are illustrated and compared using three real data sets. The simulations and empirical results overall favor both the segmented and Bayesian methods of estimation, which simultaneously estimate the change points and the other model parameters, though only the Bayesian method is able to handle both continuous and dis-continuous change point problems successfully. If it is known that regression lines are continuous then the segmented method ranked first among methods.

MSC:

62J05 Linear regression; mixed models
62F15 Bayesian inference
62F10 Point estimation
65C40 Numerical analysis or methods applied to Markov chains
65C60 Computational problems in statistics (MSC2010)

Software:

Segmented
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] Andrews, D.W.K.: Tests for parameter instability and structural change with unknown change point. Econometrica 61, 821–856 (1993) · Zbl 0795.62012 · doi:10.2307/2951764
[2] Andrews, D.W.K., Ploberger, W.: Optimal tests when a nuisance parameter is present only under the alternative. Econometrica 62, 1383–1414 (1994) · Zbl 0815.62033 · doi:10.2307/2951753
[3] Bacon, D.W., Watts, D.G.: Estimating the transition between two intersecting straight lines. Biometrika 58, 525–534 (1971) · Zbl 0224.62035 · doi:10.1093/biomet/58.3.525
[4] Bai, J.: Estimation of a change point in multiple regressions models. Rev. Econ. Stat. 79, 551–563 (1997) · doi:10.1162/003465397557132
[5] Carlin, B.P., Gelfand, A.E., Smith, A.F.M.: Hierarchical Bayesian analysis of change point problems. Appl. Stat. 41, 389–405 (1992) · Zbl 0825.62408 · doi:10.2307/2347570
[6] Chen, C.W.S., Lee, J.C.: Bayesian inference of threshold autoregressive models. J. Time Ser. Anal. 16, 483–492 (1995) · Zbl 0833.62083 · doi:10.1111/j.1467-9892.1995.tb00248.x
[7] Chen, C.W.S., Gerlach, R., Lin, A.M.H.: Falling and explosive, dormant and rising markets via multiple-regime financial time series models. Appl. Stoch. Models Bus. Ind. 26, 28–49 (2010). · Zbl 1224.91185 · doi:10.1002/asmb.765
[8] Chib, S.: Bayes regression with autoregressive errors: a Gibbs sampling approach. J. Econom. 58, 275–294 (1993) · Zbl 0775.62068 · doi:10.1016/0304-4076(93)90046-8
[9] Chow, G.: Tests of equality between sets of coefficients in two linear regressions. Econometrica 28, 591–605 (1960) · Zbl 0099.14304 · doi:10.2307/1910133
[10] Fearnhead, P.: Exact and efficient Bayesian inference for multiple changepoint problems. Stat. Comput. 16, 203–213 (2006) · doi:10.1007/s11222-006-8450-8
[11] Ferreira, P.E.: A Bayesian analysis of a switching regression model: a known number of regimes. J. Am. Stat. Assoc. 70, 370–374 (1975) · Zbl 0319.62046 · doi:10.1080/01621459.1975.10479875
[12] Hastings, W.K.: Monte-Carlo sampling methods using Markov chains and their applications. Biometrika 57, 97–109 (1970) · Zbl 0219.65008 · doi:10.1093/biomet/57.1.97
[13] Hinkley, D.: Bootstrap methods (with discussion). J. R. Stat. Soc., Ser. B. 50, 321–337 (1988) · Zbl 0671.62037
[14] Hinkley, D., Schechtman, E.: Conditional bootstrap methods in the mean-shift model. Biometrika 74, 85–93 (1987) · Zbl 0604.62036 · doi:10.1093/biomet/74.1.85
[15] Julious, S.A.: Inference and estimation in a changepoint regression problem. J. R. Stat. Soc. Ser. D, Stat. 50, 51–61 (2001) · doi:10.1111/1467-9884.00260
[16] Kim, H.J., Siegmund, D.: The likelihood ratio test for a changepoint in simple linear regression. Biometrika 76, 409–423 (1989) · Zbl 0676.62027 · doi:10.1093/biomet/76.3.409
[17] Lerman, P.M.: Fitting segmented regression models by grid search. Appl. Stat. 29, 77–84 (1980) · doi:10.2307/2346413
[18] Loader, C.R.: Change point estimation using nonparametric regression. Ann. Stat. 24, 1667–1678 (1996) · Zbl 0867.62033 · doi:10.1214/aos/1032298290
[19] MacNeill, I.B., Mao, Y.: Change-point analysis for mortality and morbidity rate. In: Sinha, B., Rukhin, A., Ahsanullah, M. (eds.) Applied Change Point Problems in Statistics, pp. 37–55 (1995)
[20] Metropolis, N., Rosenbluth, A.W., Rosenbluth, M.N., Teller, A.H., Teller, E.: Equations of state calculations by fast computing machines. J. Chem. Phys. 21, 1087–1091 (1953) · doi:10.1063/1.1699114
[21] Muggeo, V.M.R.: Estimating regression models with unknown break-points. Stat. Med. 22, 3055–3071 (2003) · doi:10.1002/sim.1545
[22] Muggeo, V.M.R.: Segmented: an R package to fit regression models with broken-line relationships. News. R Proj. 8(1), 20–25 (2008)
[23] Pastor, R., Guallar, E.: Use of two-segmented logistic regression to estimate change-points in epidemiologic studies. Am. J. Epidemiol. 148, 631–642 (1998) · doi:10.1093/aje/148.7.631
[24] Quandt, R.E.: The estimation of the parameters of a linear regression system obeying two separate regimes. J. Am. Stat. Assoc. 53, 873–880 (1958) · Zbl 0116.37304 · doi:10.1080/01621459.1958.10501484
[25] Quandt, R.E.: Tests of the hypotheses that a linear regression system obeys two separate regimes. J. Am. Stat. Assoc. 55, 324 (1960) · Zbl 0095.13602 · doi:10.1080/01621459.1960.10482067
[26] Seber, G.A.F., Wild, C.J.: Nonlinear Regression. Wiley, New York (1989)
[27] Smith, A.F.M., Cook, D.G.: Straight lines with a change-point: a Bayesian analysis of some renal transplant data. Appl. Stat. 29, 180–189 (1980) · Zbl 0432.62076 · doi:10.2307/2986304
[28] Stephens, D.A.: Bayesian retrospective multiple-changepoint identification. Appl. Stat. 43, 159–178 (1994) · Zbl 0825.62412 · doi:10.2307/2986119
[29] Ulm, K.: A statistical method for assessing a threshold in epidemiological studies. Stat. Med. 10, 341–349 (1991) · doi:10.1002/sim.4780100306
[30] Vostrikova, L.J.: Detecting ”disorder” in multidimensional random process. Soviet Math. Dokl. 24, 55–59 (1981) · Zbl 0487.62072
[31] Zhou, H.L., Liang, K.Y.: On estimating the change point in generalized linear models. In: IMS Collections Beyond Parametrics in Interdisciplinary Research: Festschrift in Honor of Professor Pranab K. Sen, vol. 1, pp. 305–320 (2008)
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.