on markets & machine learning
Xinyue Zhang (Celina)
Systems Engineering
University of Pennsylvania
Empirical asset pricing · Financial machine learning · AI for quantitative research
I study financial markets through mathematics, empirical evidence, and machine learning, and build tools to make quantitative research more rigorous and scalable.
From my research notebook
Featured research
Deep Learning for Option Pricing
Accepted in MATHEMATICA NUMERICA SINICA
CEV & SLV · Accuracy and stability Working paperIVOL & Asset Pricing
22 forecasts · 25 years of Chinese equities
How forecast errors change inference Research softwareAlphaRefinery
LLM-assisted factor discovery and refinement
Explore the architectureAbout me
I am a first-year master’s student in Systems Engineering at the University of Pennsylvania , expecting to graduate in May 2028. I received my BSc in Applied Mathematics from the Central University of Finance and Economics (CUFE) and spent Fall 2025 as a visiting student at UC Berkeley. I am currently a Research Assistant at the Wharton School.
My research interests are empirical asset pricing, financial machine learning, and AI for quantitative research . My AI experience spans deep learning for financial time series, attention-based models, model ensembles, and LLM-assisted research workflows. I also contributed to a multi-agent futures trading project in Professor Jian Li’s research group , exploring the application of AI agents to quantitative trading. On the quantitative side, I have worked on alpha-factor discovery, empirical asset pricing, portfolio construction, and strategy evaluation. I am excited to bring these areas together: using AI both to model financial data and to improve how research ideas are generated, tested, and developed.
I aspire to become a quantitative researcher , combining mathematics and programming to explore financial markets. I find markets fascinating because they bring together human behavior, uncertainty, and patterns that continually evolve. I enjoy immersing myself in a research question, learning unfamiliar methods, and working through experiments to understand what the results are telling me. As AI becomes a larger part of the research process, I want to help build tools and methods that make that process more capable and rigorous.
Through quantitative research roles at investment and proprietary trading firms, I have gained experience across the full strategy pipeline—from sourcing and preparing data to signal development, portfolio construction, backtesting, and trading implementation. I also built AlphaRefinery , an LLM-assisted platform for structured alpha-factor research.
I am interested in research collaborations and quantitative research internship opportunities.
Away from the screen
Beyond research
Different coastlines, new perspectives, and a little adventure.