curriculum vitae

University of Maryland, College Park – M.S. in Computer Science

• 5^th year M.S. in Computer Science

University of Maryland, College Park – B.S. in Computer Science

• Major GPA: 3.84/4.0
• Dean's Honors List
• Machine Learning Specialization
• Selected Coursework: Guarantees and Analyses of Machine Learning Algorithms (Ph.D. level course), Mechanism Design for Social Good (Ph.D. level course), Intro. to Machine Learning, Intro. to Artificial Intelligence, Computer Vision, Natural Language Processing, Advanced Data Structures, Design and Analysis of Algorithms, Numerical Methods.

University of Maryland, College Park – B.S. in Mathematics

• Major GPA: 3.71/4.0
• Dean's Honors List
• Statistics Specialization

Graph Visualization at Scale with Kyrix

• Research Project at UMD, advised by Dr. Leilani Battle, ...
• work in progress...

Exploring Transferability of Fairness in Machine Learning

• Graduate Course Project at UMD, advised by Dr. Furong Huang
• Investigated the transfer of fairness across domains when there is a presence of a learnable latent variable in the data.
• Generated synthetic data representing varying levels of H∆H-divergence, conditioned on sensitive attributes and an artificial binary latent variable, to test the transferability of fairness.
• Implemented Gaussian Process Latent-Variable model (TensorFlow) to investigate the transfer of fairness using synthetic data to verify our hypothesis.

Guided Human-in-the-Loop Hyperparameter Tuning

• Research Project at UMD, advised by Dr. Leilani Battle and Dr. Furong Huang
• Developed an interactive visualization system with two students to guide the hyperparameter tuning processes
• Leveraged decision nodes from random forests to estimate tight bounds for hyperparameter ranges using Python, and presented these ranges in a parallel plot using D3.js
• Created an importance metric for hyperparameters using functional-ANOVA
• Evaluated our system using A/B testing in a small user study with 6 participants

Data Scientist Intern at Kantar

• Independent project that investigated user personality classification from text to assist recommender systems and customer support.
• Implemented text classification models using Python through classical methods (e.g. Bag-of-Words, SVM, Naïve Bayes, Random Forest) and advanced methods (BERT and FastText).
• Presented results of classification methods to and potential classification metrics to an internal team

Summer Intern at BMW Autonomous Drive

• Used Docker to set up an open-source platform named EvalAI to test internal machine learning models.
• Created an automated back-up system for a containerized Postgres database