City-Scale Bayesian Spatial Modeling for House Pricing
Shirley Ren, Emily B. Fox

Citation
Shirley Ren, Emily B. Fox. "City-Scale Bayesian Spatial Modeling for House Pricing". Talk or presentation, October, 2015; Poster presented at the 2015 TerraSwarm Annual Meeting.

Abstract
In our applications of interest, we are faced with a large collection of time series that individually only provide a few observations. To make this concrete, we consider an application of modeling house sales prices from a set of geographical regions, e.g. census tracts. Existing methods for constructing housing indices are computed at a coarse spatial granularity, such as metropolitan regions. This coarse granularity does not have the representative power to encode the fine price dynamics apparent in local markets, such as neighborhoods and census tracts, and therefore leads to distorted price predictions. A challenge in moving to estimates at, for example, the census tract level is the sparsity of spatiotemporally localized house sales observations. Our work addresses the data sparsity challenge by leveraging observations from multiple local geographic regions (e.g., census tracts) discovered to have correlated dynamics. We propose a Bayesian nonparametric approach to identify an unknown number of clusters of correlated trends. We explore methods for scalability and parallelizability of computations. Our initial analysis treats the geographic unit of interest as a census tract and examines all house sales in the Seattle metropolitan area from 1997 to 2013. We then examine methods to define the geographic unit itself, rather than using pre-defined census tract regions. Instead of assuming that neighborhoods are defined by Euclidean distance, we propose an optimization based graph algorithm to discover neighborhoods of houses that have similar attributes and are closely connected by roads. Our discovered regions are at a finer scale than census tracts, and even in this case our methods described above produce a house index at this hyperlocal neighborhood level, with better predictive performance as compared to the index at the census tract level.

Electronic downloads

Internal. This publication has been marked by the author for TerraSwarm-only distribution, so electronic downloads are not available without logging in.

Citation formats

• HTML

  Shirley Ren, Emily B. Fox. <a
  href="http://www.terraswarm.org/pubs/652.html"><i>City-Scale
  Bayesian Spatial Modeling for House
  Pricing</i></a>, Talk or presentation,  October,
  2015; Poster presented at the <a
  href="http://terraswarm.org/conferences/15/annual"
  >2015 TerraSwarm Annual Meeting</a>.

• Plain text

  Shirley Ren, Emily B. Fox. "City-Scale Bayesian Spatial
  Modeling for House Pricing". Talk or presentation, 
  October, 2015; Poster presented at the <a
  href="http://terraswarm.org/conferences/15/annual"
  >2015 TerraSwarm Annual Meeting</a>.

• BibTeX

  @presentation{RenFox15_CityScaleBayesianSpatialModelingForHousePricing,
      author = {Shirley Ren and Emily B. Fox},
      title = {City-Scale Bayesian Spatial Modeling for House
                Pricing},
      month = {October},
      year = {2015},
      note = {Poster presented at the <a
                href="http://terraswarm.org/conferences/15/annual"
                >2015 TerraSwarm Annual Meeting</a>.},
      abstract = {In our applications of interest, we are faced with
                a large collection of time series that
                individually only provide a few observations. To
                make this concrete, we consider an application of
                modeling house sales prices from a set of
                geographical regions, e.g. census tracts. Existing
                methods for constructing housing indices are
                computed at a coarse spatial granularity, such as
                metropolitan regions. This coarse granularity does
                not have the representative power to encode the
                fine price dynamics apparent in local markets,
                such as neighborhoods and census tracts, and
                therefore leads to distorted price predictions. A
                challenge in moving to estimates at, for example,
                the census tract level is the sparsity of
                spatiotemporally localized house sales
                observations. Our work addresses the data sparsity
                challenge by leveraging observations from multiple
                local geographic regions (e.g., census tracts)
                discovered to have correlated dynamics. We propose
                a Bayesian nonparametric approach to identify an
                unknown number of clusters of correlated trends.
                We explore methods for scalability and
                parallelizability of computations. Our initial
                analysis treats the geographic unit of interest as
                a census tract and examines all house sales in the
                Seattle metropolitan area from 1997 to 2013. We
                then examine methods to define the geographic unit
                itself, rather than using pre-defined census tract
                regions. Instead of assuming that neighborhoods
                are defined by Euclidean distance, we propose an
                optimization based graph algorithm to discover
                neighborhoods of houses that have similar
                attributes and are closely connected by roads. Our
                discovered regions are at a finer scale than
                census tracts, and even in this case our methods
                described above produce a house index at this
                hyperlocal neighborhood level, with better
                predictive performance as compared to the index at
                the census tract level.},
      URL = {http://terraswarm.org/pubs/652.html}
  }
  

Posted by Emily B. Fox on 9 Oct 2015.
Groups: services