Yukai L.

Key Skills

Software

Other

Top Subject Matter

Finance

Statistics

Machine Learning

Top Data Types

Text

Document

Top Task Types

Data Collection

Computer Programming/Coding

Freelancer Overview

I have experience working with quantitative financial data, machine learning pipelines, and large-scale data analysis projects that closely align with AI training and data evaluation workflows. During my quantitative research internships at Tianhong Asset Management and CCB Principal Asset Management, I worked extensively with structured financial datasets, feature engineering, factor evaluation, model validation, and statistical analysis using Python, pandas, NumPy, scikit-learn, and PyTorch. I developed and evaluated quantitative factors using Information Coefficient (IC), Sharpe ratio, turnover, and cross-sectional testing, and built GRU-based time-series prediction models for A-share equities using over a decade of market data. These projects required careful data cleaning, labeling, quality control, anomaly detection, and interpretation of model outputs. In addition to quantitative finance, I have research experience involving machine learning, clustering, regression analysis, and time-series forecasting. I co-authored a peer-reviewed publication on climate risk assessment and have worked on projects involving K-means clustering, ARIMA forecasting, Random Forest models, and statistical validation. My background in mathematics, statistics, and risk engineering enables me to work effectively on AI training data tasks involving numerical reasoning, coding evaluation, analytical annotation, data verification, and model output assessment. I am detail-oriented, comfortable reviewing complex information, and experienced in maintaining consistency and accuracy in data-driven workflows. Education includes Master of Science, Duke University (2027) and Bachelor of Science, University of North Carolina at Chapel Hill (2025).

Labeling Experience

Financial Time-Series Data Labeling and Quantitative Model Evaluation

Computer Code ProgrammingData Collection

Worked on large-scale financial time-series datasets for quantitative research and machine learning applications. Performed data cleaning, preprocessing, feature engineering, and validation on A-share equity market data covering over 10 years of historical trading records. Constructed and evaluated hundreds of quantitative features using statistical metrics such as Information Coefficient (IC), ICIR, Sharpe ratio, turnover, and cross-sectional ranking performance. Prepared structured training datasets for GRU-based predictive models, including forward-return labeling, feature normalization, anomaly handling, and consistency checks. Conducted model evaluation and quality validation to ensure reliability of predictive outputs. Tasks involved numerical reasoning, structured data annotation, model output assessment, and quality assurance for large-scale machine learning workflows.

2026 - 2026

Education

Work History