This paper analyzes results from automated valuation models (AVMs) produced by academics and professionals. A large, well-documented database was prepared and made available to anyone wishing to join the research. An IAAO project permitted nationally known model builders (CAMA and AVM) as well as local and state/province professionals to demonstrate the current state of the art within the appraisal profession. This paper reports the results for both academic and professional models from the IAAO extension effort. The focus is on out-of-sample predictive power. The spatial and temporal features of the models are compared in order to learn about those techniques likely to improve predictive performance. Thus, this paper seeks to establish and document best practice CAMA models.
Data processing in assessment, Mass appraisal techniques
Clapp, J. M., & O'Connor, P. M. (2008). Automated valuation models of time and space: Best practice. Journal of Property Tax Assessment & Administration, 5(2), 57-67. Retrieved from https://researchexchange.iaao.org/jptaa/vol5/iss2/6