Yanfei Zhou

Welcome to my website. I am Yanfei Zhou (周砚菲), a final-year Ph.D. Candidate (expected graduation May, 2026) in Statistics at the Department of Data Sciences and Operations (DSO) of the University of Southern California (USC), Marshall School of Business. At USC, I am working with Professor Matteo Sesia on Conformal Inference and Deep Learning.

Before joining USC Marshall, I completed my Master’s degree in Statistics at the University of Chicago and my Bachelor’s degree in Statistics with Finance at the London School of Economics (LSE).

My current research focuses on Uncertainty Quantification (UQ) for machine learning and deep learning model predictions. Specifically, we approach this via conformal prediction, a framework that converts point predictions into reliable and informative prediction sets or intervals. Centered on conformal prediction, our research develops new methods that improve the model training process and address various issues such as fairness and data privacy. These methods can be applied to a variety of application domains such as robot motion planning, financial stock forecasting, or image recognitions. Additionally, I have begun working on large language models, focusing on uncertainty-aware model evaluation.

News

Oct 08, 2025	I am on the industry job market looking for Research Scientist, Applied Scientist, Data Scientist, or Quant Researcher positions starting in May 2026!
Aug 01, 2025	Completed Data Scientist Summer Internship atMicrosoft.
Jun 01, 2025	Received the Marshall PhD Fellowship (three awardees per department per year).
Jan 15, 2025	Received the Best Poster Award at the SEEDS 2025 Conference held by the Marshall DSO Statistics Group. Details: SEEDS Conference 2025.
Jan 10, 2025	Received the 2025 University Outstanding Teaching Assistant Award (three awardees per university per year).

Publications

2024

NeurIPS

Conformal Classification with Equalized Coverage for Adaptively Selected Groups

Yanfei Zhou, and Matteo Sesia

In Advances in Neural Information Processing Systems (NeurIPS), 2024

📄 Paper Summary Code Poster

This paper introduces a conformal inference method to evaluate uncertainty in classification by generating prediction sets with valid coverage conditional on adaptively chosen features. These features are carefully selected to reflect potential model limitations or biases. This can be useful to find a practical compromise between efficiency—by providing informative predictions—and algorithmic fairness—by ensuring equalized coverage for the most sensitive groups. We demonstrate the validity and effectiveness of this method on simulated and real data sets.
ICML

Conformalized Adaptive Forecasting of Heterogeneous Trajectories

Yanfei Zhou, Lars Lindemann, and Matteo Sesia

In International Conference on Machine Learning (ICML), 2024

📄 Paper Summary Code Poster

We present a new conformal method for generating simultaneous forecasting bands guaranteed to cover the entire path of a new random trajectory with sufficiently high probability. Prompted by the need for dependable uncertainty estimates in motion planning applications where the behavior of diverse objects may be more or less unpredictable, we blend different techniques from online conformal prediction of single and multiple time series and ideas for addressing heteroscedasticity in regression. This solution is both principled, providing precise finite-sample guarantees, and effective, often leading to more informative predictions than prior methods.

2023

ICML

Conformal inference is (almost) free for neural networks trained with early stopping

Ziyi Liang , Yanfei Zhou, and Matteo Sesia

In International Conference on Machine Learning (ICML), 2023

📄 Paper Summary Code Poster

We propose conformalized early stopping, a method that combines early stopping with conformal calibration while efficiently reusing a single hold-out set. This yields models that are accurate and provide reliable uncertainty quantification without multiple data splits. We develop practical implementations for outlier detection, multiclass classification, and regression.

2022

NeurIPS

Training uncertainty-aware classifiers with conformalized deep learning

Bat-Sheva Einbinder, Yaniv Romano, Matteo Sesia , and Yanfei Zhou

In Advances in Neural Information Processing Systems (NeurIPS), 2022

📄 Paper Summary Code Poster

We develop a training strategy for deep multiclass classification that improves model calibration, leading to more informative, reliable uncertainty estimates. Its core is a novel loss function that mitigates overfitting and overconfidence often observed with cross-entropy training. We show the method’s reliability and efficiency on synthetic and real datasets such as CIFAR-10.