Machine Learning Asset Management

Ying Han Dai

*****@*******.*** New York State cell: 949-***-****

EDUCATION

Cornell University, Ithaca, NY December 2024

Master of Engineering in Financial Engineering. GPA 3.7 Selected Coursework: Machine Learning With Large Data Set, Data mining and Machine learning; Reinforcement Learning; Stochastic Calculus; Extreme Values in Finance, Asset Management, Crypto and BlockChain, Market Making. Boston University, Boston, MA May 2023

Bachelor of Arts in Mathematics and Computer Science, Minor in Economics. Dean's list in all Semesters; Magna Cum Laude GPA 3.86 TECHNICAL SKILLS

Python, Excel, PowerPoint, R, SQL, LaTeX, Excel, Data Processing, Machine Learning, Back testing. EXPERIENCE

Quantitative Research Intern, HuaTai Security, Analyst Team, Shenzhen, China. Apr 2024 - Sept 2024

● Researched,summarized and reported most recent papers on fields such as factor model, LLM application in finance, etc

● Built a pipeline that systematically backfilled trading data; Optimized the pipeline to query and fetch data with lower overhead.

● Create features/factors from mid-frequency orderbook such as probability weighted liquidity levels. Forecasted realized volatility using these features across a variety of products.

● Implemented and trained different models (XGBoost, Random Forest, LightGBM, etc) after standardization to combine factors, results showed improvement of RankIC by average 20% along with different weighting methods between models.

● Designed and implemented a rolling multi-factor quantitative stock selection model to carry out the research of stock selection strategy.

Analyst Intern, IG Wealth, Toronto, ON. June 2024 - Aug 2024

● Developed and implemented Hierarchical Risk Parity model for portfolio management, achieving a 10% reduction in max- drawdown and 2% to 3% increase in annual return compared to baseline models.

● Implemented and analyzed a premium rate investment strategy in A-share and SEHK markets; utilizing monthly lowest premium stocks achieved an annual return rate of 17.5% with max drawdown of 19% between 2005 till present.

● Researched and analyzed quantitative signals across the crypto market. Fetching and cleaning data, feature generation based on prior market observations, and modeling to generate forecasts.

● Responsible for presenting investment portfolio performances during client meetings with data visualization. PROJECTS

Data And Strategy Analysis For Midpoint Order Volumes, MEMX, NYC, NY Aug 2024 - Current

● Conducted a comprehensive analysis of existing midpoint trading strategies, including classification methods and fill-rate filters.

● Designed a data processing pipeline and optimized the markout calculation process; improving time efficiency by 35%.

● Developed an "efficient frontier" model illustrating the trade-off between different fill-rate and markout.

● Classified stock into multi-sections by performing distribution analysis on spread and trade types. Designed and discovered optimal fill-rate signals for each individual stock section by doing correlation analysis with markouts. Market Forecasting, Cornell University, New York City, NY Oct 2024 - Dec 2024

● Conducted Exploratory Data Analysis to visualize feature relationships, identify skewness, and detect separation boundaries.

● Engineered and transformed features to address outliers, enhancing model robustness; utilized various techniques ranging from PCA to regularization to reduce dimensionality and mitigate noise.

● Designed and implemented multiple machine learning models, achieving an 88% accuracy rate for order execution predictions; developed a specialized forecasting model for Next Mid prices. Long-Short Portfolio Strategy, Cornell University, Optimal Portfolio Strategies, NYC, NY Sept 2024 - Oct 2024

● Designed a portfolio investment strategy to maximize stock selection impact while accounting for transaction and shorting costs

● Generated factors covariance matrices and then applied PCA to reduce dimensionality, addressing collinearity among risk factors.

● Designed pipelines for data preprocessing that successfully handled all mismatched data across multiple files.

● Implemented the Markowitz Optimization model to leverage the most significant factors while controlling for noise in stock alpha estimates. The overall strategy achieved a 15% annual return and a Sharpe ratio of 1.75 ACTIVITIES/INTERESTS & CERTIFICATES

Fitness Personal Trainer: Part-time fitness trainer (guided over 10 clients) and men's physique competitor. Equity Trading: Self -developed and back-tested trading strategies; combined strategies are estimated to achieve 20% annual return.

Contact this candidate