Abstract:
|
New York City park managers work with limited information when responding to natural disasters. They base decisions largely on institutional knowledge, finding it difficult to incorporate the big data generated from residents using the City's 311 system to report damaged and downed trees. Even worse, when used, analysts interpret call patterns as reflective of need without accounting for variability in individuals' underlying propensity to use 311. We are working with City agencies to investigate alternative, highly-interpretable models for predicting the propensity of individuals to call 311 in the hours and days following major storms. The result will leverage socio-demographic information from the decennial census and a New York-specific "tree census," to help the City predict and understand the relationship between the number of requests for tree service in a given census tract and the actual need. This work will allow public managers to make better decisions by incorporating 311 information into the decision of where to send work crews after the next storm.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.