A Data Enthusiast's Portfolio
Portfolio
Toxic Behavior Classifier in Gaming
Keywords: Web Scraping, NLP, LSTM, BERT
Cats and Dogs Classification
Compared different combinations of feature extraction methods and machine learning models
to classify images of cats and dogs.
Keywords: Random Forest, SVM, CNN, XGBoost
Bike Share Demand Forecasting
Forecasted hourly bike share demand after selecting and engineering
features.
Keywords: Time Series, Random Forest, XGBoost
Recommender Systems
Developed model-based and memory-based recommender systems
using implicit and explicit voting datasets.
Keywords: Collaborative Filtering
Horror Authors Comparison
Studied similarities and differences among horror authors by
performing exploratory data analysis techniques
such as sentiment and tf-idf analyses.
Keywords: NLP, Sentiment Analysis, Tf-Idf Analysis
Market Liquidity Resilience Forecasting
Forecasted market resilience of stocks following large trades by looking
at bid and ask prices.
Keywords: Time Series, Lasso Regression, GBM
About Me
After working at Morgan Stanley for two years,
I'm now a Master of Engineering Student at UC Berkeley studying
Operations Research. I'm currently working with Glooko
to predict worsening symptoms for patients with diabetes
and with GGWP to identify toxic behavior in gaming communities.
I enjoy telling data-driven stories by creating pretty visualizations and weaving compelling narratives.
As a lifetime student, I'm always on the lookout for challenging projects!
I'm currently looking for Data Scientist and Machine Learning Roles that start after I graduate in May 2021.
If you would like to reach out to talk about
any of my projects or experience, I'd be more than happy to chat!