From wanting to use AI for social good to an IEEE conference acceptance
Nitya built a machine learning model that predicts housing insecurity from socioeconomic proxy indicators, giving social services an early-warning tool before families lose their homes. Her Random Forest classifier reached 82.5% accuracy in flagging at-risk individuals. The paper was accepted at IEEE ICPSP 2025.
AI for Social Good
Accepted at IEEE ICPSP 2025
IEEE ICPSP 2025, 2025
Nitya Kaki, YRI Fellow
A high school student passionate about social justice who wanted to use AI for social good but did not know where to start.
IEEE conference accepted with a housing insecurity prediction model at 82.5% accuracy.
“My Random Forest model predicts housing insecurity with 82.5% accuracy. YRI mentorship helped me get accepted to IEEE ICPSP 2025.”
Accepted at the IEEE International Conference on Psychology and Social Policies 2025
Random Forest model reaching 82.5% accuracy in predicting housing insecurity risk
Identified socioeconomic proxy indicators serving as early warning signs
Neil Voore, Elsevier journal submission