Bayesian statistics provides a principled framework for probabilistic modeling, uncertainty quantification, and decision-making, making it a powerful tool for modern data science and machine learning. At its core, Bayesian methodology is about updating beliefs in light of new evidence, making it particularly well-suited for handling uncertainty in real-world data problems. Bayesian approaches enable robust modeling and principled inference, integrating prior knowledge with observed data to provide interpretable and coherent results. 

Bayesian inference is widely used across various domains, including healthcare, finance, artificial intelligence, and industry, where data-driven decision-makingis critical. Recent advances in Bayesian computation, such as scalable Markov chain Monte Carlo (MCMC), variational inference, and Bayesian deep learning, have significantly expanded the applicability of Bayesian methods in large-scale and complex data problems. These developments make Bayesian statistics an essential component of modern data science.

This session aims to bring together experts working at the intersection of Bayesian statistics and data science, showcasing recent methodological advances and practical applications. We aim to high-light the impact of Bayesian inference on machine learning and demonstrate how these methods can be applied to real-world problems.

Program

Applications in Data Science (two talks and one Q&A, 30 minutes)

• Harrison Bo Hua Zhu, Assistant Professor at the Section of Health Data Science and AI, Department of Public Health, University of Copenhagen. His research interests include probabilistic machine learning, deep learning methods, and their applications in infectious disease modeling and phylogenetics.
  
  
• Anna Menacher, medical innovation scientist at Novo Nordisk in Denmark, working on causal inference and medical applications. During her PhD in Statistics at the University of Oxford, she focused on scalable Bayesian models and variable selection through sparsity priors.

Bayesian Methodology (two talks and one Q&A, 30 minutes)

• Charlie Chao Zhang, Post-Doctoral scientist in the Department of Applied Mathematics and Computer Science at DTU. His research interests include Bayesian inverse problems and computational methods.
  
  
• Déborah Sulem, Assistant Professor in the Faculty of Informatics at the Università della Svizzera italiana. Her research interests include networks, point processes, Bayesian inference, high-dimensional statistics, and explainability in machine learning.

Bayesian Computation (two talks and one Q&A, 30 minutes)

• Zhihao Wang, Ph.D. researcher at the Department of Mathematical Sciences, University of Copenhagen. His research focuses on MCMC, Bayesian inference, and Markov processes for data science applications.
  
  
• Kamélia Daudel, Assistant Professor of Statistics at ESSEC Business School, Paris. Her research focuses on approximate Bayesian inference, particularly variational inference methods that go beyond parametric variational distributions.
  
  

Targeted audience

This mini-symposium is intended for a broad audience, ranging from Bayesian theorists to applied data scientists. We aim to present recent advances in Bayesian methodology and their integration with machine learning, showcasing both theoretical and applied perspectives. 

The session will demonstrate how Bayesian methods can be leveraged in real-world data science problems, such as healthcare analytics, finance, and AI interpretability. By bringing together researchers and practitioners, we hope to facilitate discussions on how Bayesian approaches shape the future of data science and empower non-experts to apply Bayesian techniques in their work.

Organizers

Jun Yang (main organizer), Assistant Professor of Statistics in the Department of Mathematical Sciences at the University of Copenhagen.