A collection of projects and papers I completed for various courses while completing a Bachelor of Arts in Computer Science, a Master of Science in Computational Science and Engineering, and a Doctor of Philosophy in Computer Science. Subjects span computer science, statistics, applied mathematics, chemistry, physics, signal processing, medicine, and more!
-
Developing a token-level explainability method to analyze social biases generated by text-to-image models (e.g. Stable Diffusion). (Spring 2023)
Course Title: Interpretability and Explainability in Machine Learning (CS 282br)
Collaborator(s): Lucas Monteiro Paes, Sree Harsha Tanneru
Tools: Python (PyTorch, Diffusers, Transformers) -
A dynamic precision selection scheme for quantizing neural network weights to reduce inference energy costs. (Fall 2021)
Course Title: Computing at Scale (CS 242)
Collaborator(s): Coleman Hooper, James Kim, Jiho Lee
Tools: Python (PyTorch) -
A deep learning system (CNN + LSTM) for using satellite imagery to predict the time and location of fires in Madagascar. (Spring 2021, Link)
Course Title: AI for Social Impact (CS 288)
Collaborator(s): Manana Hakobyan, Jessica Edwards
Acknolwedgements: This project was mentored by Professor Christopher Golden.
Tools: Python (PyTorch, FastAI, Hydra) -
GPU-accelerated MRI segmentation via spectral clustering. (Fall 2020, Link)
Course Title: Spectral Graph Theory (CS 229r)
Tools: Python (PyTorch) -
Accelerating Bayesian compressed sensing for multi-contrast MRI reconstruction. (Fall 2020, Link)
Course Title: Data Acquisition and Image Reconstruction in MRI (MIT 6.556)
Acknolwedgements: This project was mentored by Professor Demba Ba and Professor Berkin Bilgic.
Tools: Python (PyTorch) -
Optimization-based approaches for enforcing fairness in machine learning. (Spring 2019, Link)
Course Title: Advanced Optimization (APMTH 221)
Collaborator(s): Amil Merchant '19
Tools: Python (PyTorch) -
An open-source platform for evaluating causal inference models on common datasets. (Spring 2019, Link)
Course Title: Advanced Topics in Machine Learning (CS 282r)
Collaborator(s): Amil Merchant '19, Suproteem Sarkar '19
Acknolwedgements: This project was mentored by Alexander D'Amour during his time at Google Brain.
Tools: Python (NumPy, SciKit-Learn) -
An implementation of the Hartree-Fock algorithm for automatically calculating interesting properties of small molecules from first principles. (Fall 2018, Link)
Course Title: Quantum and Statistical Foundations of Chemistry (PS 10)
Acknolwedgements: This project was mentored by Michael Mavros.
Tools: Python (NumPy, SciPy) -
Interactive visualizations with storytelling for American gun violence data. (Fall 2018, Link)
Course Title: Data Visualization (CS 171)
Collaborator(s): Emily Chen '19, Michel Li '19, Daniel Seong '19
Tools: JavaScript (D3), HTML/CSS -
Examining various attack and defense methods for neural networks with respect to adversarial examples. (Spring 2018, Link)
Course Title: Computational Science and Engineering Capstone Project Course (AC 297)
Collaborator(s): ZiZi Zhang '18, Thomas Chang '19, Amil Merchant '19
Acknolwedgements: This project was mentored by Ekin Dogus Cubuk during his time at Google Brain, as well as by Patrick Ohiomoba, Eleni Kaxiras, and Professor Pavlos Protopapas.
Tools: Python (Keras, TensorFlow) -
Unaligned style transfer of sentence sentiment using Adversarially Regularized Autoencoders (ARAEs). (Spring 2018, Link)
Course Title: Machine Learning for Natural Language (CS 287)
Collaborator(s): Melissa Yu '19
Tools: Python (PyTorch) -
Fair and interpretable scoring rule models for the credit scoring problem. (Spring 2018, Link)
Course Title: Topics at the Interface between Computer Science and Economics (CS 236)
Collaborator(s): Melissa Yu '19
Tools: Python (NumPy), CPLEX -
Developing an EM algorithm for clustering time series in applications such as epilepsy recognition and neuronal firing rate analysis. (Fall 2017)
Course Title: Advanced Machine Learning (CS 281)
Acknolwedgements: This project was mentored by Professor Demba Ba and Yingzhuo (Diana) Zhang.
Tools: Python (NumPy) -
Clustering biometric data of preterm births using hierarchical Bayesian mixture models. (Spring 2017)
Course Title: Advanced Scientific Computing: Stochastic Methods for Data Analysis, Inference and Optimization (APMTH 207)
Tools: Python (PyMC3, Theano) -
Various practically-oriented projects for applying machine learning to the real world, including molecular fingerprinting, malware detection, music recommendation, and reinforcement learning for AI-driven game playing. (Spring 2017, Link)
Course Title: Machine Learning (CS 181)
Collaborator(s): Matthew Huang '19, Melissa Yu '19
Tools: Python (NumPy, Pandas, SciKit-Learn, PyGame) -
An analysis of different predictive modeling techniques (e.g. random forest, generalized linear model, artifical neural network) for determining whether or not a Colorado citizen voted in the 2016 election. (Spring 2017, Link)
Course Title: Statistical Sleuthing Through Generalized Linear Models (Stat 149)
Collaborator(s): Matthew Huang '19, Melissa Yu '19, Phillip Yu '19
Tools: R, Python (NumPy, SciKit-Learn) -
Evaluating and comparing the relative efficacies of different optimization algorithms (e.g. SGD, Adam, L-BFGS, CG) in training neural networks. (Fall 2016, Link)
Course Title: Advanced Scientific Computing: Numerical Methods (APMTH 205)
Collaborator(s): Robert Chen '17
Tools: Python (TensorFlow) -
Martingale-based algorithms for detecting various types of anomalies in time series. (Fall 2016, Link)
Course Title: Introduction to Data Science (CS 109a)
Collaborator(s): Melissa Yu '19
Tools: Python (NumPy, SciKit-Learn) -
Designing and implementing linear/integer programs for real-world optimization problems in cancer radiation treatment and paired kidney exchange. (Fall 2016)
Course Title: Introduction to Optimization: Models and Methods (APMTH 121)
Collaborator(s): Ana Carell '17, Karly Zlatic '17, Matthew Huang '19, Melissa Yu '19
Tools: AMPL, CPLEX, MATLAB -
An OCaml-based application that utilizes recursive algorithms to randomly generate mazes and compute the solutions to them. (Spring 2016, Link)
Course Title: Abstraction and Design in Computation (CS 51)
Collaborator(s): Melissa Yu '19
Tools: OCaml -
A linear model for predicting MLB batter salary from season statistics. (Spring 2016)
Course Title: Statistical Sleuthing Through Linear Models (Stat 139)
Collaborator(s): Salvador Barragan-Santana '18, Aaron Sachs '19
Tools: R -
Budgitt: A dynamic website based off the MVC architecture that helps users keep track of their budget over regular spending periods. (Fall 2015, Link)
Course Title: Introduction to Computer Science (CS 50)
Collaborator(s): Joy Jin '19, Helen Wu '19
Tools: PHP, JavaScript, MySQL, HTML/CSS