As a passionate data scientist with a robust background in math & economics, I bring a unique blend of academic rigor and practical experience to my projects. I specialize in reducing complex phenomena to their essential components, using careful study design and statistical analyses to precisely address the question at hand. I aim to be constantly growing, refining my technical skills and challenging my beliefs through dialog and data.
For my capstone project at the General Assembly Data Science Immersive, I developed new modeling techniques to predict the outcomes of players' early games on online chess platform Lichess.org. My model uses chess engine evaluations and neural networks to achieve an additional 20% accuracy improvement upon Lichess's state-of-the-art model, and the techniques I employed can be used for better probabilistic predictions and matchmaking for new players.
A group project with Stanley Azuakola and Atigon Hongchumpol from the General Assembly Data Science Immersive. We developed a model that takes the geographic coordinates of a village in Africa and predicts whether that village suffers from extreme poverty. Our model uses publicly available satellite imagery along with convolutional neural networks and ensembling to catch 79% of cases of extreme poverty while maintaining an overall accuracy of 85%.
I used ensemble models and elemantary NLP techniques to classify potential music software customers as "composers" or "producers" based on the messages they posted on Reddit. My model correctly classifies 93% of potential customers.
I used simple linear regression techniques to predict the sale prices of homes in Ames, Iowa. Through careful data cleaning, encoding of categorical variables, and feature selection, I created models that are both highly predictive and easily interpretable. My best model was the winner of a General Assembly Kaggle Competition.
A joint paper with Brian Jabarian. We designed a 4-treatment incentivized experiment with more than 700 participants to test hypotheses about how people behave when facing uncertainty and ambiguity. We showed that a majority of people don't behave according to widely-used models of decision-making, and we used statistical methods to test various hypotheses explaining such behavior.
Princeton University
MA in Economics
University of Chicago
BA with Honors in Mathematics
Data Scientist
NCRI & Narravance
Data Science Fellow
General Assembly
Teaching Assistant
Princeton University
•Games of Strategy, Chance & Deduction
•Causal Analysis
•Classical Music
•Russian Literature
•Cooking