About Me

Greetings 🌍!

I’m Boyu Fan, a senior at UC Berkeley, where I’m double majoring in Economics and Data Science with a minor in Statistics. Outside academics, you can find me playing tennis, trying out new recipes, or playing pieces by Vivaldi and Schubert.

🔍 Research Interest

High-Dimensional Data Analysis, Bayesian Hierarchical Models for Longitudinal Data, Causal Inference in Observational Studies, Uncertainty Quantification, Functional Data Analysis, Human AI Interface, Applications in Interpretable Machine Learning

🌱 Conservationist

In the heart of Kenya, I found my calling as a conservationist with the Lion Light project, a initiative blending technology and environmental stewardship.

As the Chief Advocate, I’ve been instrumental in installing innovative lion lights, which uses blinking light patterns to ward off lions from the villager’s livestock. My role extends beyond just installation; it involves research, community engagement, and raising awareness. By raising $7000, we’ve expanded the project, installing 800 additional lights and earning heartfelt thanks from the local Kenyan community.

More information can be found at my specific portfolio entry.

🔬 Research Experience

Lawrence Berkeley National Laboratory | Research Assistant (September 2023 - Present)

Fortunate to be advised by Alex Sim, K. John Wu, and Jinoh Kim, I have been employing advanced deep learning methodologies, notably Long Short-Term Memory (LSTM) and Conditional Recurrent Neural Networks (RNN), to analyze time series data using TensorFlow. This work involves conducting detailed analyses of Tstat format logs to develop innovative predictive methodologies and tools, primarily aimed at estimating and improving network performance.

Haas School of Business, Berkeley Operations and Behavioral Analytics Lab | Research Assistant (September 2023 - Present)

Fortunate to be advised by Park Sinchaisri, my research centers on the dynamics of Human-AI collaboration, specifically through the lens of sequential decision-making and inverse reinforcement learning. A key aspect of our work involves the development of a two-player game employing Markov Decision Processes (MDP) and Reinforcement Learning (RL). This project is designed to explore the role of communication between players in collaborative settings. Our focus is on formulating strategic tips provided during the game and assessing their impact on team collaboration and decision-making. The game serves as a model for studying and enhancing Human-AI interactions, emphasizing the balance of team and individual goals through effective communication strategies.

📊 Selected Work Experience

Deloitte | Risk Advisory Modeling Intern (May - August 2023)

In my role at Deloitte, I focused on enhancing sovereign risk models by performing rigorous data cleaning and feature extraction on datasets from the World Bank and IMF. I employed advanced feature engineering techniques like K-means and agglomerative clustering, along with univariate analysis using tools such as Somer’s D, ROC, and information value. My work culminated in the construction of a sovereign risk classification model using logistic regression, decision trees, and random forests, significantly improving the model’s accuracy from 40% to 78% and streamlining data processing with proficient use of SQL.

Apple | Data Analysis Engineering Intern (July - August 2022)

At Apple, I collaborated with the Global Channel sales team to develop and execute effective business and sales strategies. My key contributions included orchestrating the design and implementation of A/B tests, applying one-tail hypothesis testing to understand customer behavior, leading to an 18% increase in unique viewer metrics. I also optimized sales strategies using a reinforcement model and skillfully analyzed sales reports from distributors, employing advanced data engineering techniques to yield insightful business strategies.