Start Date
6-4-2018 1:15 PM
End Date
6-4-2018 2:15 PM
Description
All models used by Philadelphia OPA employ spatial attributes. This presentation shows how we measure and transform spatial attributes and use them in our models. Demonstrations include how spatial attributes relate to other location variables in providing explanatory within the model, and measures of effectiveness. The presentation also includes identification of 12 candidate attributes and tests for the best measurement of each. The most effective measurement was determined, and data was transformed for use in both linear and non-linear valuation models. Results of this research and testing resulted in the decision by Philadelphia Office of Property Assessment to include spatial attributes in all valuation models. We are currently in our third major revaluation project using these models, which have created major improvements to the quality of our assessments that will be described to the audience and can be used globally.
Publication Date
April 2018
Recommended Citation
Keene, Kevin and Williams, James, "Transformation and spatial attributes in regression modeling" (2018). International Research Symposium. 20.
https://researchexchange.iaao.org/irs/irs18/sessions/20
Transformation and spatial attributes in regression modeling
All models used by Philadelphia OPA employ spatial attributes. This presentation shows how we measure and transform spatial attributes and use them in our models. Demonstrations include how spatial attributes relate to other location variables in providing explanatory within the model, and measures of effectiveness. The presentation also includes identification of 12 candidate attributes and tests for the best measurement of each. The most effective measurement was determined, and data was transformed for use in both linear and non-linear valuation models. Results of this research and testing resulted in the decision by Philadelphia Office of Property Assessment to include spatial attributes in all valuation models. We are currently in our third major revaluation project using these models, which have created major improvements to the quality of our assessments that will be described to the audience and can be used globally.