CV
Education
- Ph.D in Computer Science, Georgia Institute of Technology, 2028 (Expected), GPA: 4.0
- MS in Machine Learning, Carnegie Mellon University, 2023 (Left to Pursue Ph.D.), QPA: 4.0
- B.A. in Computer Science (Magna Cum Laude) and Mathematics, Cornell University, 2022, GPA: 3.91.
Research Experience
- Graduate Researcher 08/2023~Present
- Conducting research on fusing generative modeling approaches for applications to data assimilation in physical systems
- Developed a latent assimilation algorithm which works well for high-dimensional bayesian filtering problems
- Worked on integrating method with latent dynamics in a subsequent paper for fast simulation
- Undergraduate Researcher (with Professor. Volodymyr Kuleshov) 09/2020~05/2023
- Conducted research in the area of aleatoric and epistemic uncertainty estimation for calibrated uncertainty prediction using quantile loss functions
- Demonstrated that by integrating quantile loss with autoregressive flows, we can remove the need for the
- Jacobian while retaining sampling and uncertainty estimation capabilities
- Extended research to a non-autoregressive flow architecture for faster training and sampling
- Natural Language Processing Research Intern at Cornell Tech 05/2019~08/2019
- Conducted NLP research under Professor Yoav Artzi at Language in Context Lab
- Developed system for using collaborative interactions to study second language acquisition in goal-specified environment with Unity and Python
- Used NewsRoom dataset to generate articles with multiple summary labels for parameter based summary generation with various Python architectures
- Statistics Research Intern at Hong Kong University of Science and Technology 06/2017~10/2017
- Research mathematics and statistics under Professor Yuan Yao
- Completed a pairwise comparison machine learning algorithm from cancer cell and drug sensitivity data using R
- Model placed third (highest rank was second) on Kaggle leaderboard among graduate and undergraduate students)
- Enhanced model from database on cancerrxgene.org and conducted detailed analysis
Clubs And Organizations
- Intelligent Systems Team Lead at Cornell Data Science 09/2019~05/2022
- Snapbee, using Deep Learning to identify handwritten text and figures in general documents (specifically testing and modifying the DeepFigures framework to work for handwritten inputs)
- Pleio Project, using causal inference to obtain information about when and how to best remind patients to take their medicine
- Work on combining Bayesian Networks with Kernel Density Estimators to reduce the effects of high-dimensional data
- Using Reinforcement Learning to develop a Pokerbot agent
- Organized and planned team meetings, held reading groups, and oversaw project progress through biweekly standups
Industry Experience
- Machine Learning Research Intern at CardioPhi 03/2023~08/2023
- Constructed Transformer and ResNet based models to process ECGs for detection and prediction of heart arrhythmia
- Collaborated with front-end engineers to deliver information for an insightful UI
- Wrote technical portions of grants for NSF and NIH funding
- Worked on paper ECGBERT: Understanding the Hidden Language of ECGs
- Software Engineering Intern at Adobe Inc. 05/2021~08/2021
- Improved cost and efficiency of a generative search algorithm by incorporating native speedups using C++ and lower-level code as part of an innovation team developing cutting-edge and currently realizable technology
- Cut out entirety of heavy cloud-computing cost while retaining similar time performance against cloud GPUs
- Software Engineering Intern at Adobe Inc. 05/2020~08/2020
- Researched and developed a ML pipeline in Python for efficiently resolving problems and directing users to right resource
- Boosted performance compared to pre-existing Elasticsearch approach from 30% to 90% accuracy, from a hand-generated validation dataset
- Deployed the internal tool through an online docker container and Amazon Lambda (serverless), then integrated with a slack bot as a slash command
- Machine Learning Intern at Vital Scientist INC. 06/2018~07/2018
- Used detailed scrum and sprint to collaborate with the project team, conducted a presentation regarding our final product, a (locally hosted) website which recommended restaurants based on previous likes
- Extracted data with SparkSQL and analyzed a restaurant model based on Yelp dataset with Numpy, using stochastic processes to decrease calculation complexity of the collaborative filtering algorithm
Publications
