| 

| 

IMDB Movie Modeling
IMDB Movie Modeling

A Data Analysis Project Using the IMDB Dataset on Kaggle. Goal is build model for log movie profitability (log revenue/cost). Features include native dataset features, network features from running PageRank on actor co-occurances, and text features generated by training word vectors on movie descriptions. Final model was Randrom Forest with Feature Selection using Lasso CV, R2 obtained was 0.647.

python
US-China Venture Capital Social Graph
US-China Venture Capital Social Graph

As part of my Master's Thesis, I scraped around 4 years of Chinese venture capital investment data from a public source and compared it to venture capital investment data in the United States. I was interested in whether the rate of co-investment of established early-stage VC firms with less established early stage firms was higher in China or the US (answer, the former). This became quite a heavily involved data procurement, cleaning, management, and analysis project.

PythonBeautifulSoupMongoDB