Yuteng Chen Email & Phone Number

Boston, Massachusetts, United States

• Implemented a gradient-boosted decision tree model to address numerical errors in cyber exposure data and designed a simulated anomaly dataset for model tuning and evaluation, which automates the error detection process with a 70% true positive rate and 95% true negative rate with real-time prediction capability.• Developed a machine-learning solution to separate mixed-up retention and attachment point values in UK cyber exposure data, utilizing transfer learning to build two models, a primary/excess classifier and a retention predictor, that achieved a combined R-squared score of 0.84.• Constructed a statistical model to estimate the entire insurance tower size from limited data, overcoming missing data challenges and providing accurate estimates across all revenue buckets and industries.• Built a revenue imputation method after EDA and feature selection to address the missing revenue data problem, reducing the RMSE from 2.4 to 1.6 compared to the previous method.

Shanghai, China

• Utilized Python to scrape product information and customer reviews of Midea air conditioners from JD.com. Applied NLP techniques to analyze customer sentiments, evaluating Midea's strengths and weaknesses in the domestic air conditioner market based on user feedback on aspects such as appearance, cooling performance, installation, etc.• Transcribed audio files of meetings involving company employees and industry experts into text using OpenAI's LLM API. Extracted and summarized key information from the transcriptions with improved accuracy and level of detail in essential information extraction through prompt engineering.• Conducted market and due diligence research on leading companies in the commercial aerospace and wind power sectors to compile preliminary reports.

Jinhua, Zhejiang, China

• Monitored and tracked financial status using Python to analyze time-series data of debtor companies. Built a model to predict potential defaults based on historical financial data and media sentiment analysis, assisting in early identification of reputation risks.• Extracted and aggregated financial data for the department’s weekly reports, creating visualizations using Tableau to provide actionable insights for senior management.• Conducted data analysis on bond underwriting transactions by collecting and cleaning financial datasets from issuing companies, focusing on key metrics such as debt ratios, financial health, and bond performance.

Master Of Science - Ms, Data Science, 4.0

Bachelor Of Science - Bs, Applied Mathematics, Economics, 3.9, Summa Cum Laude

Activities and Societies: Phi Beta Kappa