Applied Data Scientist
Princeton, Nj
Primary Technologies Used: Python (Jupyter Lab, Pandas, Sci-kit learn, Seaborn, Matplotlib)• Analyzed and delivered key business understanding of electronic consent data across multiple functional domains from ~4000 patients across Schizophrenic, ADPKD, Bipolar, and MDD clinical trials. o Performed data sourcing, data pre-processing, data integration, and association analyses while interfacing with various business and clinical stakeholders.o Results were used to improved electronic consent rate by 20%.• Conducted data pre-processing and exploratory analyses of patient actigraphy data and adverse event data (~1000 patients, Epilepsy).o Performed feature engineering to define the “normal” activity of patients.o Developed an Elastic Net Linear Regression model to predict patient normal activity with a mean absolute percent error of 15%. This model was used to generate additional data for various other clinical trial analyses.• Researched and identified socio-economic and environmental factors that affect longitudinal patient clinical scores. Current work is in process of being submitted for journal publications.o Sourced various internal and external datasets (US Census, historical weather data, Air Quality Index, Clinical Trial Data).o Processed and integrated environmental data with existing clinical data using geo-analysis (~1000 patient records).• Utilized data-driven approaches to manage UX/UI design and write functional requirement specifications for innovative digital products.o Launched products: REVA, MITRAM (available on iOS App and Android Play Store).o Performed A/B testing and product analytics to improve MITRAM application survey response rate by 15%.