Data Matters is the weekly seminar series of the Data Science Initiative. Talks are held Wednesdays during the academic year, 4–5:30 pm, at 164 Angell Street, 3rd floor.
The Data Science Initiative includes scholars with backgrounds in the physical, biological, computational, and social sciences. Data Matters is intended to stimulate conversations and collaboration by bringing multiple perspectives to challenging data-driven problems and talks are structured to be more of an interactive experience than traditional academic seminars. The format is generally a 25-30 minute presentation followed by 45 minutes of interaction with the audience, mediated by a discussant.
Applications of Social Network Analysis and Agent-Based Modeling to Study the Population Health Impacts of US Carceral Systems
Aditya Khanna, Assistant Professor, Behavioral and Social Sciences
My talk will open with an overview of the scope and utility of social network analysis and agent-based modeling in public health. We will then consider two recent application projects. The first is on HIV transmission among Black men who have sex with men (MSM) in Chicago, which is home to Cook County Jail, the largest single-site jail system in the US. The talk will describe the development of an agent-based network modeling (ABNM) framework to estimate the impact of CCJ on the Black MSM community and their social networks. We will show how addressing the needs of this community is vital to achieving aims of the Getting to Zero HIV Elimination initiative (https://gtzillinois.hiv).
The second part of the talk will describe a project that concerns the Rhode Island criminal-legal system. We will look at an ongoing modeling effort to understand the impact of incarceration on smoking and alcohol use disorder in the social networks of persons involved with the correctional system in the state of Rhode Island. We will examine the network impacts of the “syndemic” of incarceration, smoking, and alcohol use, and consider how modeling can inform biobehavioral interventions and policy work in this area. The project builds on the Repast4Py toolkit for agent-based modeling (https://repast.github.io/repast4py.site/index.html) and the EMEWS (Extreme-scale Model Exploration with Swift) platforms for large-scale model simulation and analysis (https://emews.github.io).
My work is largely supported by the Brown University Center for Addiction and Disease Risk Exacerbation (CADRE, P20 GM 130414), and uses HPC resources provided by the Center for Computing and Visualization.
Light refreshments will be available. This talk will be recorded.