On Spatio-Temporal Model with Diverging Number of Thresholds and its Applications in Housing Market

Abstract

Spatio-temporal data analysis is an emerging research area due to the development and application of novel computational techniques allowing for the analysis of large spatio-temporal databases. We consider a general class of spatio-temporal linear models, where the number of structural breaks can tend to infinity. A procedure for simultaneously detecting all the change points is developed rigorously via the construction of adaptive group lasso penalty. Consistency of the multiple change point estimation is established under mild technical conditions even when the true number of change points

s_n

\documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}\begin{document}$$s_n$$\end{document}

diverges with the series length n. The adaptive group lasso can be substituted by the group lasso and other non-convex group selection penalty functions such as group SCAD or group MCP. The simulation studies demonstrate that our procedure is stable and accurate. Two empirical examples from property market, including the housing transaction price in Baton Rouge and the commodity apartment price in Hong Kong, are analyzed to fully illustrate the proposed methodology.

Keywords

Change points / Balanced panel data dynamic linear models / Group selection / Real estate market / Spatio-temporal data

References

[1] Anselin, L.: Spatial Econometrics: Methods and Models, 4. Springer Science & Business Media (1988)

[2] AueA, HorváthL. Structural breaks in time series. J. Time Ser. Anal., 2013, 3411-16

[3] BaiJ. Estimating multiple breaks one at a time. Economet. Theor., 1997, 133315-352

[4] BaiJ. Estimation of a change point in multiple regression models. Rev. Econ. Stat., 1997, 794551-563

[5] BaiJ. Common breaks in means and variances for panel data. J. Econ., 2010, 157178-92

[6] BaiJ, PerronP. Computation and analysis of multiple structural change models. J. Appl. Economet., 2003, 1811-22

[7] BelL, Bar-HenA, PetitR, CheddadiR. Spatio-temporal functional regression on paleoecological data. J. Appl. Stat., 2011, 384695-704

[8] BernardiMS, SangalliLM, MazzaG, RamsayJO. A penalized regression model for spatial functional data with application to the analysis of the production of waste in venice province. Stoch. Env. Res. Risk Assess., 2017, 31123-38

[9] Bleakley, K., Vert, JP.: The group fused lasso for multiple change-point detection. arXiv preprint arXiv:1106.4199 (2011)

[10] BlundellR, BondS. Initial conditions and moment restrictions in dynamic panel data models. J. Econ., 1998, 871115-143

[11] BrehenyP, HuangJ. Group descent algorithms for nonconvex penalized linear and logistic regression models with grouped predictors. Stat. Comput., 2015, 252173-187

[12] BrooksC, TsolacosSReal estate modelling and forecasting, 2010CambridgeCambridge University Press

[13] CliffA, OrdJSpatial autocorrelation, 1973LondonPion

[14] De WachterS, TzavalisE. Detection of structural breaks in linear dynamic panel data models. Comput. Stat. Data Anal., 2012, 56113020-3034

[15] ElhorstJP. Specification and estimation of spatial panel data models. Int. Reg. Sci. Rev., 2003, 263244-268

[16] Elhorst, JP.: Spatial econometrics: from cross-sectional data to spatial panels. Springer Science & Business Media (2013)

[17] FanJ, LiR. Variable selection via nonconcave penalized likelihood and its oracle properties. J. Am. Stat. Assoc., 2001, 964561348-1360

[18] FanJ, HallP, MartinM, PatilP. Adaptation to high spatial inhomogeneity using wavelet methods. Stat. Sin., 1999, 9: 85-102

[19] FanJ, FengY, WuY, et al. . Network exploration via the adaptive lasso and scad penalties. The Ann. Appl. Stat., 2009, 32521-541

[20] FearnheadP, VasileiouD. Bayesian analysis of isochores. J. Am. Stat. Assoc., 2009, 104485132-141

[21] FinleyA, SangH, BanerjeeS, GelfandA. Improving the performance of predictive process modeling for large datasets. Comput. Stat. Data Anal., 2009, 53: 2873-2884

[22] FrickK, MunkA, SielingH. Multiscale change point inference. J. Royal Stat. Soc.: Series B (Statistical Methodology), 2014, 763495-580

[23] FryzlewiczP. Wild binary segmentation for multiple change-point detection. Ann. Stat., 2014, 4262243-2281

[24] GiraldoR, Dabo-NiangS, MartinezS. Statistical modeling of spatial big data: An approach from a functional data analysis perspective. Stat. Probab. Lett., 2018, 136: 126-129

[25] HarchaouiZ, Lévy-LeducC. Multiple change-point estimation with a total variation penalty. J. Am. Stat. Assoc., 2010, 1054921480-1493

[26] Hepşen, A., Vatansever, M.: Forecasting future trends in dubai housing market by using box-jenkins autoregressive integrated moving average. Int. J. Housing Markets Anal. (2011)

[27] HuangB, WuB, BarryM. Geographically and temporally weighted regression for modeling spatio-temporal variation in house prices. Int. J. Geogr. Inf. Sci., 2010, 243383-401

[28] HuangH, CressieN. Spatio-temporal prediction of snow water equivalent using the kalman filter. Comput. Stat. Data Anal., 1996, 22: 159-175

[29] JandhyalaV, FotopoulosS, MacNeillI, LiuP. Inference for single and multiple change-points in time series. J. Time Ser. Anal., 2013, 344423-446

[30] Jin, B., Shi, X., Wu, Y.: A novel and fast methodology for simultaneous multiple structural break estimation and variable selection for nonstationary time series models. Statistics and Computing 1–11 (2013)

[31] JinB, WuY, ShiX. Consistent two-stage multiple change-point detection in linear models. Can. J. Stat., 2016, 442161-179

[32] JinX, CarlinB. Multivariate parametric spatio-temporal models for county level breast cancer survival data. Lifetime Data Anal., 2003, 11: 5-27

[33] JosephL, WolfsonDB. Estimation in multi-path change-point problems. Commun. Stat.-Theory Methods, 1992, 214897-913

[34] JosephL, WolfsonDB. Maximum likelihood estimation in the multi-path change-point problem. Ann. Inst. Stat. Math., 1993, 453511-530

[35] KeY, LiJ, ZhangW. Structure identification in panel data analysis. Ann. Stat., 2016, 44: 1193-1233

[36] KimD. Common breaks in time trends for large panel data with a factor structure. Economet. J., 2014, 173301-337

[37] LeeDJ, ZhuZ, ToscasP. Spatio-temporal functional data analysis for wireless sensor networks data. Environmetrics, 2015, 265354-362

[38] LeeLF, YuJ. Some recent developments in spatial panel data models. Reg. Sci. Urban Econ., 2010, 405255-271

[39] LeSageJP, PaceRK. Models for spatially dependent missing data. The J. Real Estate Fin. Econ., 2004, 292233-254

[40] Li, CW.: Spatial autocorrelation and liquidity in hong kong’s real estate market. HKU Theses Online (HKUTO) (2010)

[41] LiJ, JinB. Multi-threshold accelerated failure time model. Ann. Stat., 2018, 46: 2657-2682

[42] Niu, YS., Hao, N., Zhang, H.: Multiple change-point detection: a selective overview. Statistical Science 611–623 (2016)

[43] PaceRK, BarryR, GilleyOW, SirmansC. A method for spatial-temporal forecasting with an application to real estate prices. Int. J. Forecast., 2000, 162229-246

[44] PerronP, et al. . Dealing with structural breaks. Palgrave Handbook of Econ., 2006, 12278-352

[45] QuZ, PerronP. Estimating and testing structural changes in multivariate regressions. Econometrica, 2007, 752459-502

[46] Ramsay, JO., Ramsay, T., Sangalli, LM.: Spatial functional data analysis. In: Recent advances in functional data analysis and related topics, Springer, 269–275 (2011)

[47] RenQ, BanerjeeS, FinleyA, HodgesJ. Variational bayesian methods for spatial data analysis. Comput. Stat. Data Anal., 2011, 55: 3197-3217

[48] RodriguesG, PrangleD, SissonS. Recalibration: a post-processing method for approximate bayesian computation. Comput. Stat. Data Anal., 2011, 126: 53-66

[49] Smith, RL., Kolenikov, S., Cox, LH.: Spatiotemporal modeling of pm2. 5 data with missing values. J. Geophys. Res.: Atmos. 108(D24) (2003)

[50] WangH, LengC. A note on adaptive group lasso. Comput. Stat. Data Anal., 2008, 52125277-5286

[51] WestT, WorthingtonAC. Macroeconomic risk factors in australian commercial real estate, listed property trust and property sector stock returns. J. Financ. Manag. Prop. Constr., 2006, 112105

[52] YangY, ZouH. A fast unified algorithm for solving group-lasso penalize learning problems. Stat. Comput., 2015, 2561129-1141

[53] YanoskyJD, PaciorekCJ, LadenF, HartJE, PuettRC, LiaoD, SuhHH. Spatio-temporal modeling of particulate air pollution in the conterminous united states using geographic and meteorological predictors. Environ. Health, 2014, 13163

[54] Yao, YC., Au, ST.: Least-squares estimation of a step function. Sankhyā: The Indian J. Stati., Series A 370–381 (1989)

[55] YuanM, LinY. Model selection and estimation in regression with grouped variables. J. Royal Stat. Soc.: Series B (Statistical Methodology), 2006, 68149-67

[56] ZhangCH, et al. . Nearly unbiased variable selection under minimax concave penalty. Ann. Stat., 2010, 382894-942

Funding

National Natural Science Foundation of China(No.11571337)

Key Programme(No.71631006)

Natural Sciences and Engineering Research Council of Canada(Grant No. RGPIN-2017-05720)

RIGHTS & PERMISSIONS

School of Mathematical Sciences, University of Science and Technology of China and Springer-Verlag GmbH Germany, part of Springer Nature