Shan Zhong

Work on autonomous vehicle dispatch algorithm.

• Teach three sections of STAT 201 Lab, 72 people each. Wrote blogs style statistical learning tutorials. • Completed three large-scale projects: S&P 500 Stock Price Prediction Using Technical, Fundamental and Text Data. , Functional Clustering of Fictional Narratives Using Vonnegut Curves. and Multi-agent value and policy learning to build prediction intervals for massive correlated stock data and construct portfolios under risk-measure. • Dissertation: Deep Learning, Clustering, and Decision Process Approaches for Modeling Time Series Data

• Automated web data scraping from 91 life insurance companies for over 20,000 Critical Illness, health, and accident insurance products via python Scrapy. • Use regular expressions and TF-IDF word frequency language model to extract keywords from documents and construct databases. • Product valuation in terms of net present value (NPV), construct recommendation system

• Developed Poisson regression models to predict actual claim count, with past 30 years of data. Calculated adjusted loss ratio volatility and correlation between 15 different lines of business, for 1400 companies. • Drafted weekly research reports to present to chief actuary Mr.Steve White. • Used SQL language to manipulate raw data and construct Access databases.

Evaluated winning strategy for different 52-deck poker games and designed new game variants for use in gaming devices. • Used excel VBA Monte Carlo decision tree simulations to validate game programs. • Design algorithm, and calculate the odds ratio according to the weight change of possible outcomes.

• Used three different databases, to build models to assess the family assets of incoming freshmen based on their home addresses filled for over 10,000 students. • Utilized Excel’s Vlookup function to input and clean data. • Grade other students’ exams and assignments

• Used excel to process about 70,000 rows of premium income data of around 200,000 people, filtered and categorized the data. • Calculated and analyzed accident insurance claims data from multiple regions. • Processed hundreds of accident insurance underwriting documents, review and seal.

Dissertation: Deep Learning, Clustering, and Decision Process Approaches for Modeling Time Series Data