THIS PART OF THE WEBPAGE IS UNDER DEVELOPMENT AND WILL BE UPDATED SOON
Analyzed medical data to estimate variance after using cross-validation and imputation techniques in classification problems. My poster showcasing this research won the Best Poster Award, 3rd place, at the Student Poster Competition at the Statistical Practice conference.
In this project, we investigated how HIV-1 and its proteins contribute to neuroinflammation leading to HIV-associated neurocognitive disorders (HAND). We identified IFIT3 and STAT1 as key genes upregulated during infection, even under antiretroviral therapy. To analyze neuronal and brain tissue images, we applied convolutional neural networks (CNNs), which classified protein expression patterns with near-perfect accuracy. By integrating machine learning, we validated our biological findings and highlighted IFIT3 as a potential therapeutic target.
Collaborated with Kwaai AI Lab to design and implement homomorphic encryption methods for privacy-preserving vector search, co-developing two novel algorithms in the Mathematical Problems in Industry Workshop.
Built linear regression and principal component analysis models to estimate CO2, CH4 fluxes in Brazilian fish ponds.
Feature engineering to characterize high water level events using hydrological, meteorological, and tidal processes.