Xinyu(Katherine) He

Generated features using minute-level cryptocurrency data and employed RankIC and LassoCV for feature selection. Subsequently, applied an LSTM model to analyze time-series data from cryptocurrencies' features to make predictions. Additionally, evaluated the trading strategies to assess a hit ratio of over 69% and annual sharp values of 0.002.

Database Design and SQL teaching assistant

Introduction to Python and Data Science teaching assistant

Deep Learning for Prediction of Business Outcomes teaching assistant

Mi Pharma is considering entering the anti-doping market. Conducted a comprehensive SWOT analysis and performed antidoping market analysis by reaching out to relevant stakeholders and leveraging data from the United States Anti-Doping Agency. Additionally, executed pricing analysis by studying competitors to show potential pricing models. Provided recommendation against Mi Pharma's entry into the anti-doping market based on the comprehensive findings.

• Conducted climate risk analysis on collateral data and Fannie Mae DUS bond utilizing Python geospatial visualization to create recommendations for investment decision-making. • Performed credit risk review on American Express National Bank, leveraging four years of data from S&P Capital IQ to evaluate financial stability, analyzed and assessed potential risks. • Analyzed MIAC pricing models for Commercial Real Estate loans and implemented a Random Forest Regression Model to identify and select most influential features for pricing.

• Cleaned and summarized primary bond market data using Wind Database and Excel, published data on company website, presented findings to management • Provided solutions to missing data to enhance financial accuracy and support the company’s risk assessment efforts