Uncovering Hidden Spatial Patterns by Hidden Markov Model

  • Ruihong Huang
  • Christina Kennedy
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5266)


Many spatial data mining and spatial modeling approaches use Euclidean distance in modeling spatial dependence. Although meaningful and convenient, Euclidean distance has weaknesses. These include providing an over simplified representation of spatial dependence, being limited to certain spatial pattern and symmetrical relationships, being unable to account for cross-class dependencies, and unable to work with categorical especially multinomial data. This paper introduces Hidden Markov Model (HMM) as an attractive approach to uncovering hidden spatial patterns. The HMM assumes that a hidden state (factor or process) generates observable symbols (indicators). This doubly embedded stochastic approach uncovers hidden states based on observed symbol sequences using two integrated sets of probabilities, transition probability and emission probability. As an alternative to Euclidean distance based approaches, the HMM measures spatial dependency by transition probabilities and cross-class correlation better capturing geographic context. HMM works with data of any measurement scale and dimension. To demonstrate the method, we assume urban spatial structure as a hidden spatial factor underlying single family housing unit prices in Milwaukee, Wisconsin, we then use the HMM to uncover four hidden spatial states from home sale prices.


GIS Hidden Markov Model spatial modeling data mining 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Aycard, O., Mari, J.F., Washington, R.: Learning to automatically detect features for mobile robots using second-order Hidden Markov Models. International Journal of Advanced Robotic Systems 1(4), 233–250 (2004)Google Scholar
  2. 2.
    Basu, S.: Analysis of spatial autocorrelation in house prices. Journal of Real Estate Finance and Economics 17(1), 61–85 (1998)CrossRefGoogle Scholar
  3. 3.
    Bierkens, M.F.P., Burrough, P.A.: The indicator approach to categorical soil data: I, Theory. Journal of Soil Science 44, 361–368 (1993)Google Scholar
  4. 4.
    Brunsdon, C., Fotheringham, A.S., Charlton, M.E.: Geographically weighted regression: a method for exploring spatial nonstationarity. Geographical Analysis 28, 281–298 (1996)Google Scholar
  5. 5.
    Can, A.: Specification and estimation of hedonic housing price models. Regional Science and Urban Economics 22, 453–474 (1992)CrossRefGoogle Scholar
  6. 6.
    Cheshire, P., Sheppard, S.: On the price of land and the value of amenities. Economica 62, 247–267 (1995)CrossRefGoogle Scholar
  7. 7.
    Cliff, A.D., Haggett, P.: On complex geographic space: Computing frameworks for spatial diffusion processes. In: Longley, P.A., Brooks, S.M., McDonnell, R., MacMillan, B. (eds.) Geocomputation: A Primer, pp. 231–256. Wiley, New York (1998)Google Scholar
  8. 8.
    Dempster, A.P., Laird, N.M., Rubin, D.B.: Maximum likelihood from incomplete data via the EM algorithm. Journal of the Royal Statistical Society, Series B 39, 1–38 (1997)MathSciNetGoogle Scholar
  9. 9.
    Dubin, R.A.: Spatial autocorrelation and neighborhood quality. Regional Science and Urban Economics 22, 433–452 (1992)CrossRefGoogle Scholar
  10. 10.
    Elfeki, A., Dekking, M.: A Markov chain model for subsurface characterization: theory and applications. Mathematical Geology 33(5), 569–589 (2001)zbMATHCrossRefMathSciNetGoogle Scholar
  11. 11.
    Ester, M., Kriegel, H.P., Sander, J.: Algorithms and applications for spatial data mining. In: Miller, H.J., Han, J. (eds.) Geographic Data Mining and Knowledge Discovery, pp. 160–187. Taylor & Francis, London (2001)Google Scholar
  12. 12.
    Evans, A.W.: The economics of residential location. Macmillan, London (1973)Google Scholar
  13. 13.
    Fong, E., Shibuya, K.: The spatial separation of the poor in Canadian cities. Demography 37(4), 449–459 (2000)CrossRefGoogle Scholar
  14. 14.
    Getis, A., Ord, J.: The analysis of spatial association by use of distance statistics. Geographical Analysis 24, 189–206 (1992)Google Scholar
  15. 15.
    Goovaerts, P.: Stochastic simulation of categorical variables using a classification algorithm and simulated annealing. Mathematical Geology 28, 909–921 (1996)CrossRefGoogle Scholar
  16. 16.
    Grether, D.M., Mieszkowski, P.: The effects of nonresidential land uses on the prices of adjacent housing: some estimates of proximity effects. Journal of Urban Economics 8, 1–15 (1980)CrossRefGoogle Scholar
  17. 17.
    Han, J., Kamber, M., Tung, A.K.H.: Spatial clustering methods in data mining. In: Miller, H.J., Han, J. (eds.) Geographic data mining and knowledge discovery. Taylor & Francis, New York (2001)Google Scholar
  18. 18.
    Hughes, J.P., Guttorp, P., Charles, S.P.: A non-homogeneous hidden Markov model for precipitation occurrence. Journal of the Royal Statistical Society (Series C): Applied Statistics 48(1), 15–30 (1999)zbMATHCrossRefGoogle Scholar
  19. 19.
    Iceland, J.: Beyond black and white: metropolitan residential segregation in multiethnic America. Social Science Research 33(2), 248–271 (2004)CrossRefGoogle Scholar
  20. 20.
    Kaufman, L., Rousseeuw, P.J.: Finding Groups in Data: An Introduction to Cluster Analysis. John Wiley & Sons, Chichester (1990)Google Scholar
  21. 21.
    Levine, M.V.: Suburban sprawl and the "secession" of the affluent: metropolitan polarization in Milwaukee: 1987-1997. University of Wisconsin-Milwaukee Center for Economic Development Policy Research Report (1999),
  22. 22.
    Levine, M.V.: Metropolitan polarization in an era of affluence: income trends in metropolitan Milwaukee since 1990. University of Wisconsin-Milwaukee Center for Economic Development Policy Research Report (2002a),
  23. 23.
    Levine, M.V.: The economic state of Milwaukee.s inner city: 1970-2000, University of Wisconsin-Milwaukee Center for Economic Development Policy Research Report (2002b),
  24. 24.
    Levine, M. V. 2003. The two Milwaukees: separate and unequal,
  25. 25.
    Li, W.: Transiogram: A spatial relationship measure for categorical data. International Journal of Geographical Information Science 20(6), 693–699 (2006)CrossRefGoogle Scholar
  26. 26.
    Li, W., Zhang, C.: A generalized Markov chain approach for conditional simulation of categorical variables from grid samples. Transactions in GIS 10(4), 651–669 (2006)CrossRefGoogle Scholar
  27. 27.
    Lovell, B.C.: Hidden Markov models for spatio-temporal pattern recognition and image segmentation. In: International Conference on Advances in Pattern Recognition - (2003)Google Scholar
  28. 28.
    Mari, J.F., Le Ber, F.: Temporal and spatial data mining with second-order hidden Markov Models. Soft Comput. 10, 406–414 (2006)CrossRefGoogle Scholar
  29. 29.
    Miller, H.J.: Tobler’s First Law and Spatial Analysis. Annals of the Association of American Geographers 94(2), 284–289 (2004)CrossRefGoogle Scholar
  30. 30.
    Miller, H.J., Wentz, E.A.: Representation and Spatial Analysis in Geographic Information Systems. Annals of the Association of American Geographers 93(3), 574–594 (2003)CrossRefGoogle Scholar
  31. 31.
    Milwaukee Neighborhood Identification Project, Milwaukee neighborhoods (2000) (Last time accessed August 17, 2007),
  32. 32.
    Movellan, J.R.: Tutorial on hidden Markov models. Machine perception laboratory online tutorials (2003) (Last time accessed May 2, 2006),
  33. 33.
    Nathan, K.S., Bellegarda, J.R., Nahamoo, D., Bellegarda, E.J.: On-line handwriting recognition using continuous parameter hiddenMarkov models. In: ICASSP 1993 (IEEE International Conference on Acoustics, Speech, and Signal Processing 1993), vol. 5, pp. 121–124 (1993)Google Scholar
  34. 34.
    Ord, J.K., Getis, A.: Local spatial autocorrelation statistics – distributional issues and an application. Geographical Analysis 27, 286–306 (1995)Google Scholar
  35. 35.
    Orford, S.: Valuing locational externalities: a GIS and multilevel modelling approach. Environment and Planning B: Planning and Design 29, 105–127 (2002)CrossRefGoogle Scholar
  36. 36.
    O’Sullivan, D., Unwin, D.J.: Geographic information analysis. John Wiley and Sons, Inc., Hoboken (2003)Google Scholar
  37. 37.
    Palay, M.G.: 1973, Facts and Figures for Community Analysis. Milwaukee Urban Observatory, University of Wisconsin-Milwaukee, Milwaukee (1970)Google Scholar
  38. 38.
    Pamuk, A.: Geography of Immigrant Clusters in Global Cities: A Case Study of San Francisco, 2000. International Journal of Urban and Regional Research 28(2), 287–307 (2004)CrossRefGoogle Scholar
  39. 39.
    Pena, J.M., Lozano, J.A., Larranaga, P.: An empirical comparison of four initialization methods for the K-Means algorithm. Pattern Recognition Letters 20, 1027–1040 (1999)CrossRefGoogle Scholar
  40. 40.
    Rast, J.: Transportation equity and access to jobs in metropolitan Milwaukee. University of Wisconsin-Milwaukee Center for Economic Development Policy Research Report (2004),
  41. 41.
    Rabiner, L.R., Wilpon, J.G., Juang, B.H.: A Segmental K-means Training Procedure for Connected with Recognition Based on Whole Word Reference Patterns. AT&T Technical Journal 65(3), 21–31 (1986)Google Scholar
  42. 42.
    Rabiner, L.R.: A tutorial on hidden Markov model and selected applications in speech recognition. Proceedings of the IEEE 77(2), 257–286 (1989)CrossRefGoogle Scholar
  43. 43.
    Robertson, A.W.S.K., Smyth, P.: Hidden Markov models for modeling daily rainfall occurrence over Brazil. Technical Report UCI-ICS 03-27, Information and Computer Science, University of California, Irvine (2003)Google Scholar
  44. 44.
    Shekhar, S., Lu, C.T., Zhang, P.: A unified approach to detecting spatial outliers. GeoInformatica 7(2), 139–166 (2003)CrossRefGoogle Scholar
  45. 45.
    Tanguay, D.O.: Hidden Markov models for gesture recognition. Thesis of Massachusetts Institute of Technology (1995)Google Scholar
  46. 46.
    Tjelmeland, H., Besag, J.: Markov random fields with higher-order interactions. Scandinavian Journal of Statistics 25, 415–433 (1998)zbMATHCrossRefMathSciNetGoogle Scholar
  47. 47.
    Tobler., W.: A computer movie simulating urban growth in the Detroit region. Economic Geography 46, 234–240 (1970)CrossRefGoogle Scholar
  48. 48.
    Varel, O.J.: Examining Patterns of Neighborhood Change in the City of Milwaukee, 1980-1990: An Ecological Approach. University of Wisconsin-Milwaukee dissertation (1999)Google Scholar
  49. 49.
    Viovy, N., Saint, G.: Hidden Markov models applied to vegetation dynamics analysis using satellite remote sensing. IEEE Transactions on Geoscience and Remote Sensing 32(4), 906–917 (1994)CrossRefGoogle Scholar
  50. 50.
    Wilkes, R., Iceland, J.: Hypersegregation in the twenty-first century. Demography 41(1), 23–36 (2004)CrossRefGoogle Scholar
  51. 51.
    Wong, D.W.S., Lee, J.: Statistical Analysis of Geographic Information. John Wiley & Sons, Inc., New Jersey (2005)Google Scholar
  52. 52.
    Zhang, J., Goodchild, M.: Uncertainty in geographic information. Taylor and Francis, New York (2002)Google Scholar
  53. 53.
    Zhang, C., Li, W.: Markov chain modeling for multinomial land-cover classes. GIScience & Remote Sensing 42(1), 1–18 (2005)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2008

Authors and Affiliations

  • Ruihong Huang
    • 1
  • Christina Kennedy
    • 1
  1. 1.Department of Geography, Planning & RecreationNorthern Arizona UniversityFlagstaffUSA

Personalised recommendations