Brief Abstract
This session will explore how machine learning and SHAP (Shapley Additive Explanations) values can be powerful tools in identifying and addressing inequities in property valuations. SHAP values, originally developed to explain complex model predictions, can be adapted to examine sales ratios and other property characteristics, helping pinpoint both vertical and horizontal inequities. This approach provides a clear understanding of which factors contribute to disparities, offering actionable insights to improve fairness in valuations. By incorporating SHAP values, assessors can not only make valuations more equitable but also save time and money by reducing the need for extensive manual analysis and rework. The transparency offered by SHAP helps identify inequities early on, preventing future issues and streamlining the assessment process. The session will cover practical examples in R and Python, guide you through interpreting results, and highlight potential challenges to be aware of. This session is aimed at professionals who want to leverage AI to make their valuations more accurate and efficient, ultimately reducing the burden of time, costs, and complications while improving equity in the process.
Start Date
3-5-2025 2:30 PM
End Date
3-5-2025 3:30 PM
Enhancing ratio studies with AI: Identifying sources of inequity with SHAP values
This session will explore how machine learning and SHAP (Shapley Additive Explanations) values can be powerful tools in identifying and addressing inequities in property valuations. SHAP values, originally developed to explain complex model predictions, can be adapted to examine sales ratios and other property characteristics, helping pinpoint both vertical and horizontal inequities. This approach provides a clear understanding of which factors contribute to disparities, offering actionable insights to improve fairness in valuations. By incorporating SHAP values, assessors can not only make valuations more equitable but also save time and money by reducing the need for extensive manual analysis and rework. The transparency offered by SHAP helps identify inequities early on, preventing future issues and streamlining the assessment process. The session will cover practical examples in R and Python, guide you through interpreting results, and highlight potential challenges to be aware of. This session is aimed at professionals who want to leverage AI to make their valuations more accurate and efficient, ultimately reducing the burden of time, costs, and complications while improving equity in the process.