The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence) Applications of Variational Autoencoders and Bayesian Optimization with José Miguel Hernández Lobato - #510
Aug 16, 2021
José Miguel Hernández Lobato, a machine learning lecturer at the University of Cambridge, shares insights on the fusion of Bayesian learning and deep learning in molecular design. He discusses innovative methods for predicting chemical reactions and explores the challenges of sample efficiency in reinforcement learning. José elaborates on deep generative models, their role in molecular property prediction, and strategies for enhancing the robustness of machine learning through invariant risk minimization. His research reveals exciting pathways in optimizing molecule discovery.
AI Snips
Chapters
Transcript
Episode notes
Bayesian Deep Learning and Uncertainty
- Bayesian deep learning offers uncertainty quantification, crucial for decision-making.
- Approximations are necessary due to the intractability of precise uncertainty estimation.
Uncertainty in Molecular Design
- Molecular design uses uncertainty to guide searches, balancing potential and uncertainty.
- Focus on molecules with good property potential and high uncertainty for faster discovery.
3D Molecular Generation
- Deep generative models learn from existing molecules to create new, synthesizable ones.
- 3D molecule generation offers advantages over graph-based methods, capturing crucial structural information.