Binhao Wang

Binhao Wang

Ph.D. student in Data Science at CityU HK. My research focuses on recommender systems, knowledge graphs and LLM.

Portfolio

Application of HSIC for Non-linear Dependency Analysis (Independent Research)

Published:

In this independent research, I applied the Hilbert-Schmidt Independence Criterion (HSIC) to assess the goodness-of-fit for complex time-series models. I demonstrated its capability to capture complex, non-linear dependencies in model residuals where traditional linear methods like Pearson correlation fail.

Knowledge Graph-based Chemical Reaction Path Finding (Undergraduate Thesis)

Published:

This project was my undergraduate thesis. I designed and implemented a data pipeline using the GPT API to perform automated semantic entity extraction from over 1,000 scientific papers, constructing a structured knowledge graph for reaction path analysis. This system improved search efficiency by 20% over traditional keyword-based baselines.