About

University of California, San Diego

I have a broad interest in machine learning, finance, software development and data science. My primary focus is on utilizing advanced quantitative methods in areas traditionally dominated by qualitative theories, aiming to bring new insights into under-explored fields across both the sciences and social sciences.

Professional Interests

Mathematics & Physics Double Major

I am interested in software development, data science, and finance, with years of on-and-off experience in the field working in web and application development. My coding skills span multiple languages, including Swift, Python, SQL, MATLAB, and React.js, among others. I'm also proficient in machine learning and data analysis techniques, using tools like Sklearn, Statsmodels, and Tensorflow. For a detailed look at my programming work, please refer to my portfolio.

Experience

Machine Learning & Applied Mathematics

Researcher | Machine Learning in Seismology & Quantitative Marine Ecology

October 2023 - Present

• Employed empirical dynamical modeling to quantitatively assess species interactions under fluctuating environmental conditions, drawing correlations with temporal population dynamics.

• Used high dimensional regression on field observations to model predictors of seismic hazards.

Researcher | Machine Learning in Financial Forecasting

March 2023 - September 2023

• Designed predictive models for forecasting macroeconomic indicators using large language models like ChatGPT and natural language processing (VADER), utilizing logistics regression, random forest and ARIMA algorithms resulting in MAPE 12% better than the baseline model.

• Built a pipeline integrating OpenAI's API, constructing, and managing large databases, and performing model validation analyses in stationarity, Pearson correlation and Granger causality.

Financial Data Engineer

October 2023 - Present

• Leveraged Oracle's API with cx_Oracle to automate and streamline financial data extraction, ensuring seamless integration and data consistency utilizing pandas and regex for data cleaning.

• Devised a reconciliation strategy using the anti-join method to find discrepancies in datasets for swift identification of unapplied payments resulting in 70% faster process than manual methods.