Dec 18, 2019
08:00 AM - 09:00 AM
Modeling Geographic Preferences for Policy Decisions
SDP Free Webinar - Co-Sponsored with the Decision Analysis Society (DAS)
Speaker: Jay Simon, Associate Professor of Information Technology & Analytics, American University.
Moderator: L. Robin Keller, Professor of Operations & Decision Technologies, UC Irvine.
Wednesday, 18 Dec - 8 am PT | 11 am ET
Geographic information systems (GIS) have become increasingly prevalent over the past 10-15 years as a tool to support policy decisions. An urban planner might want to incorporate the impact of zoning decisions on temperature and air pollution throughout a region when comparing development strategies. A state government might want to evaluate the impact of potential conservation approaches on groundwater levels throughout the state. The goal of our work is to provide a foundation for applying sound decision analysis principles to these types of decisions.
The challenge, and what differentiates decision problems based on GIS data from standard multi-attribute decisions, is that the realized level of an attribute is not a single number; rather, it is a set of levels distributed throughout a geographic region. A decision analyst must assess not only a value function and weight for each attribute, but also a set of weights for different locations. These location weights will typically capture one or more of: area, population, economic significance, or environmental sensitivity. There are straightforward conditions on preferences for that, if met, will make elicitation tractable.
After presenting the key concepts involved with geographic preferences, we will explore example applications and discuss potential ways to ease the elicitation burden on decision makers.