Title: Causal Machine Learning with Applications in Computational Biology
About the Talk: Many scientific questions are fundamentally causal in nature. Yet, existing causal inference methods are not suited for complex, high-dimensional data. Causal representation learning (CRL) fills this gap by modeling causal structure in the latent space of a machine learning model. First, I will review our work on the theoretical and algorithmic foundations of CRL across different settings. Then, I will present our current efforts on making CRL more practical and using it to address difficult out-of-distribution and extrapolation problems such as predicting the effects of unseen drug or gene perturbations from omics measurements. Causal ML requires rich experimental data, and single-cell biology offers unique opportunities for gaining new scientific insights by leveraging such principled methods.
About the Speaker: Julius is currently an AI researcher at ETH Zürich. He is also a Branco Weiss Fellow and affiliated with the ETH AI Center. His research focuses on causal inference and machine learning, particularly causal representation learning and computational biology. He completed his PhD jointly at the University of Cambridge and the Max Planck Institute for Intelligent Systems.