Skip to content
Jessica Zhu
Jessica Zhu

Jessica Zhu

Operations Research
U.S Army
Bio

Jessica Zhu is a transitioning service member and a PhD student at the University of Maryland. She previously served as an Army Captain and machine learning researcher at the Laboratory for Telecommunication Sciences. Jessica commissioned in 2017 as a Military Intelligence Officer and graduated from West Point with a degree in Operations Research and Chinese. She graduated from MIT with an SM in Operations Research through a scholarship from MIT Lincoln Laboratory in 2019. CPT Zhu also deployed an Intelligence Analyst and the Lead Data Analyst with the Combined Joint Interagency Task Force and was assigned to the Army AI Integration Center as the Director’s Executive Assistant and Technical Lead to the Intel Support to Operations Portfolio.


Model Facts: Putting AI Ethics into Practice and Establishing Trust with End Users

As machine learning enabled tools become ubiquitous in daily life, there is an increasing onus on developers to establish trust with their end users. This is especially true when creating tools that will impact historically exploited vulnerable populations. Society’s conscious and systemic biases have continued to be reflected in tools made for predictive policing, recidivism judgement, and even chat software. Building upon our previously published paper, we share our Model Facts framework. It echoes the straightforwardness of Nutrition Facts labels for users to assess the validity and biases of models without diving into technical model documentation.  This template aims to be easily extendable while decomposing accuracy and demographics into standardized and simple values. We take firearm injury research as an example of a community that will benefit from the adoption of Model Facts labels. Implementation of Model Facts will provide public health practitioners, those impacted by firearm injury, and similarly vulnerable communities greater faith in the tools models provide.