智能制造技术教育部重点实验室 学术论坛

报告题目：Introduction to Conformal Prediction: Application to Multi-label Classification

报告嘉宾：Wenge Guo Department of Mathematical Sciences New Jersey Institute of Technology, Newark, NJ, USA

报告时间：2024年7月4日下午3:00

报告地点：科技楼五楼521会议室

报告摘要：

Conformal prediction is a versatile statistical framework

that offers valid measures of uncertainty for predictions in supervised learning tasks. It ensures that prediction intervals or sets contain the true outcomes with a specified probability, thereby providing a robust measure of reliability. This methodology is particularly valuable in fields where precise uncertainty quantification is essential. In this talk, we will provide a comprehensive introduction to this interdisciplinary field, bridging machine learning and statistics. We will also explore the application of conformal prediction to multi-label classification, a complex task where each instance can be associated with multiple labels simultaneously.

嘉宾简介：

Wenge Guo is an Associate Professor of Statistics at the New Jersey Institute of Technology. He earned his PhD in Biostatistics from the University of Cincinnati and subsequently worked as a Research Fellow at the National Institute of Environmental Health Sciences for two years.

He serves on the editorial boards of three journals: Statistics & Probability Letters, PLOS ONE, and the Calcutta Statistical Association Bulletin. His primary research interests encompass Statistical Machine Learning, Multiple Testing, Selective Inference, and Statistical Methods for Clinical Trials.

欢迎工学院和学校其他院系有兴趣的师生参加！

工学院

2024年7月2日