Olivia Fan Email and Phone Number

Enhanced accuracy of core NLP Text Analytics models to above 95% in f-score, precision and recall. Optimized the model size by 86% on UNIX systems. Deployed model to production.End-to-end development of high-quality Python scripts for automated testing to ensure product functionality and downstream performance, contributed to the company’s Gerrit repository for future utilization

Used R and SQL to statistically analyze long-term behavioral data ‘Babase’ to investigate impacts of reproductive status on baboon foraging behaviorsUsed generalized estimating equations (GEEs) and generalized linear mixed models (GLMMs) to model foraging outcomes; Conducted causal inference and time series ARIMA analysis to investigate the reverse causality between network density and foraging

Analyzed health effects of chemical exposures with advanced statistical methods such as Bayesian Model Averaging (BMA), Bayesian Kernel Machine Regression (BKMR), Bayesian Regression Model (BRM), Quantile G- Computation, Weighted Quantile Sum Regression & Bayesian Weighted Sums (BWS). Sponsored by NIEHS (National Institute of Environmental Health Sciences).

Teaching Assistant for [CS 330] Design and Analysis of Algorithms, [CS 371] Elements of Machine Learning, Machine Learning and Data Mining [STA 325]Assisted professor by grading homework, holding office hours, compiling problem set solutions, and proctoring exams.

Proposed a novel adaptation of the Q Learning algorithm that combines non-linear scalarized learning updates and non-stationary action selection to learn effective policies for optimizing nonlinear welfare functionDemonstrated experimentally that our approach outperforms techniques based on linear scalarization, mixtures of optimal linear scalarizations, or stationary action selection for the Nash Social Welfare Objective.