5/16/19 | 4:15pm | E51-325
Reception to follow.
Mohammad Fazel Zirandi
The Immigration and Customs Enforcement (ICE) is responsible to enforce the immigration laws of the United States. In the past decade, various programs and guidelines have been introduced to focus ICE’s resources to deport the more dangerous and violent individuals, and to prioritize the removal of immigrants based on the severity of their crimes.
In this work, we use individual-level administrative data from ICE and the state of the art interpretable machine learning tools to answer the following questions:
1. Does the data support ICE’s claim of prioritizing deportations based on severity of crimes?
2. Given a significant change in policy in 2014, is the new policy more effective in aligning ICE’s objective of prioritizing based on crime to how the policy is actually enforced by ICE offices across the US?
3. Is there heterogeneity across ICE offices on how they adhere to the prioritization guidelines?
This work is a first step in our overall aspiration to use interpretable machine learning tools and large data to analyze and improve public policy for significant societal challenges.
This is joint work with Dimitris Bertsimas.
Bio: Mohammad Fazel‐Zarandi is a Senior Lecturer at the MIT Sloan School of Management. His research interests are in the areas of public sector operations, data-driven decision making, statistical learning, applied probabilistic modeling and statistics. Mohammad's research has been published in leading journals such as Operations Research and Management Science, and has been reported by the Los Angeles Times, the Boston Globe, the New York Times, Bloomberg, and many other major media outlets.