Arushi Sinha

Developed a predictive ML model to enhance insurance member compliance with critical Healthcare Effectiveness Data and Information Set (HEDIS) measures. Utilized advanced analytics on geocoded Social Determinants of Health (SDoH) derived from US census data for targeted interventions.

Bridged the synthetic-to-real domain gap, boosting classification model performance with controlled diffusion and Vision-Language Models.

Tutoring a group of middle school students in Mathematics as part of the STEM outreach program by the Center of Educational Outreach.

• Performed manual data extraction of log data by identifying anomalies. • Cleaned and pre-processed this data using Borehole correction charts. • Accurately estimated fluid saturation levels across the depth of oil well by log interpretation to identify the hydrocarbon zone.

• Built an algorithm to predict missing sonic logs in the Oil Industry • Trained an Ensemble learning model on a dataset of 20,525 data points, incorporating petrochemical analysis. • Achieved a root mean square error score of 3.88 for one sonic log and 9.91 for another, as validated on a separate test dataset of 3,000 data points provided by the company.

Master's in Electrical and Computer Engineering with a specialization in Machine Learning and Data Science