Short Courses

Short Course - Conformal Inference (Wednesday, January 8th, 9am-12pm) John Cherian (Stanford University) Making sense of the black box: what can conformal inference offer the theory and practice of data science?
Bio: John Cherian is a 5th-year PhD student at Stanford University supported by the Hertz Foundation. Advised by Emmanuel Candès, he works on problems in model-free inference and uncertainty quantification. He also consults for The Washington Post, where he applies this research to night-of election models. Prior to the PhD, John spent three years at D.E. Shaw Research improving molecular dynamics simulations for structural biology and drug discovery.
Abstract: New artificial intelligence models have delivered impressive performance improvements over a diverse array of tasks, but their reliability across different contexts or subpopulations is not assured. Therefore, it is important to quantify uncertainty before we take consequential actions that leverage AI outputs. This tutorial will introduce an increasingly popular framework for this task: conformal inference. Conformal methods quantify uncertainty for future predictions without making any distributional assumption whatsoever other than having i.i.d. or, more generally, exchangeable data. This talk will review the basic principles underlying conformal inference, contextualize the framework’s relationship to classical statistical methods, and survey some major contributions that have occurred in the last 2-3 years. These include (but are not limited to) new methods that enable stronger validity guarantees, further relaxation of the aforementioned distributional assumption, and trustworthy identification of the most promising subset of predictions. We will motivate each of these techniques with real-world problems, and discuss the practical obstacles that lie between the theory of conformal prediction and its practice. While the majority of the tutorial will presume an undergraduate background in probability and statistics, we will conclude this session with a more advanced discussion of the theoretical properties of conformal methods.

Short Course - GenAI 1 (Wednesday, January 8, 1pm-5pm) Ron Li (Author, Data Scientist) Master Cursor.ai: Unleash the Builder Within the Researcher
Bio: Ron Li is a data science entrepreneur and educator, with experience in both industry and academia. He graduated from USC with two master's degrees in physics and electrical engineering. Ron will be assisted by Yixiang Yao (https://www.isi.edu/directory/yixiangy/), a Computer Science Ph.D. student at the University of Southern California. Yixiang's research focuses on implementing privacy-enhancing technologies in general AI applications, including privacy-preserving entity resolution and language model privacy enhancement.
Abstract: Generative AI has redefined the boundaries of what researchers can build and achieve, turning the seemingly impossible into reality. As statisticians and mathematicians, we pride ourselves on rigorous reasoning and analytical thinking, but true innovation requires embracing the spirit of creation. Cursor.ai empowers researchers to transcend traditional limitations, enabling them to design, experiment, and innovate in ways never before imagined. In this course, we will explore two key methodological aspects of interacting with AI: how to explain complex concepts effectively and how to master efficient prompt engineering. These skills are crucial for unlocking the full potential of tools like Cursor.ai, ensuring that the AI becomes not just a tool, but a collaborator in your creative journey. Designed for master’s and beginning PhD students in statistics and mathematics, this course aims to ignite the creator’s spirit within. Through hands-on exploration of Cursor.ai’s capabilities, students will learn how to leverage AI to tackle complex problems, automate labor-intensive processes, and pioneer new approaches in research. By the end of the course, participants will emerge not just as users of technology, but as visionary creators in the era of AI-driven discovery.

Short Course - GenAI 2 (Friday, January 10th, 1pm-4pm) Haoda Fu (Amgen) Tutorial on Deep Learning and Generative AI
Bio: Dr. Haoda Fu is Head of Exploratory Biostatistics in Amgen, before that he was an Associate Vice President and an Enterprise Lead for Machine Learning, Artificial Intelli- gence, from Eli Lilly and Company. Dr. Haoda Fu is a Fellow of ASA (American Statistical Association), and IMS Fellow (Institute of Mathematical Statistics). He is also an adjunct professor of biostatistics department, Univ. of North Carolina Chapel Hill and Indiana university School of Medicine. Dr. Fu received his Ph.D. in statistics from University of Wisconsin - Madison in 2007 and joined Lilly after that. Since he joined Lilly, he is very active in statistics and data science methodology research. He has more than 100 publi- cations in the areas, such as Bayesian adaptive design, survival analysis, recurrent event modeling, personalized medicine, indirect and mixed treatment comparison, joint modeling, Bayesian decision making, and rare events analysis. In recent years, his research area focuses on machine learning and artificial intelligence. His research has been published in various top journals including JASA, JRSS-B, Biometrika, Biometrics, ACM, IEEE, JAMA, Annals of Internal Medicine etc.. He has been teaching topics of machine learning and AI in large industry conferences including teaching this topic in FDA workshop. He was board of directors for statistics organizations and program chairs, committee chairs such as ICSA, ENAR, and ASA Biopharm session. He is a COPSS Snedecor Awards committee member from 2022-2026, and also served as an associate editor for JASA theory and method from 2023, and JASA application and case study from 2025-2027.
Abstract: This course is tailored for individuals with a solid background in statistics or biostatistics, focusing on deep learning and generative AI. Starting with foundational deep learning concepts, participants will implement models using PyTorch, explore popular AI architectures like CNN, GNN, ResNet, U-net, and transformers, and address applications in medical imaging and drug discovery. The course will also introduce generative AI including GANs, VAEs, DDPM, score-based models, and also how LLM works. It offers a comprehensive insight by providing mental models into applying AI in healthcare, research, and beyond.