Doctoral student in biostatistics. Interested in spatial statistics, Bayesian statistics, and machine learning methodology as well applications in infectious disease and pandemic preparedness/surveillance.s

Implementing machine learning models in R using high-throughput multimodal automated phenotyping involving mixture models, clustering, and regression• Applying the phenotyping process (PheCAP) – running the natural language processing (NLP) tool, extracting billing data from SQL, curating the data, and running the algorithm in R – to predict rheumatoid arthritis, myocardial infarction, and acute ischemic stroke in >17,000,000 patients• Training analysts on other teams to run PheCAP• Conducting a phenome-wide association study (PheWAS) on the Interleukin-6 receptor variant using a Linux high-performance-computing cluster• Extracting:o Text information from > 1,000,000,000 EHR clinical notes with NLP tools in Pythono Genetic variants using Plink to explore the association between the UCP1 gene and obesityo Data sets in SQL in the Million Veterans Project (MVP) and the CVD Merit Project• Performing quality control on data curation and analyses run by team members• Meeting with supervisors routinely to review upcoming projects, determine analyses to perform, data to be extracted, and who to collaborate with• Organizing and maintaining project spaces, data sources and code for analyses• Presenting analysis results to biostatisticians, clinicians and project managers at weekly meetings• Reduced storage space of NLP results by 30% by introducing index-based storageIndependent work:• Developing an unsupervised method to disambiguate abbreviations in clinical notes using a random forest model, word embeddings, and expectation-maximization clustering• Applying the disambiguation model to improve phenotyping in MVP and CVD