Yee Whye Teh

1. Academic Summary



Yee Whye Teh is a leading researcher in machine learning, Bayesian statistics, and artificial intelligence, particularly known for foundational contributions to Bayesian nonparametrics, deep learning, probabilistic models, and scalable inference methods. He is recognized internationally for his theoretical and methodological work and has held academic roles in Europe and industry research leadership at DeepMind. 



2. Education


  • B.Math (Computer Science & Pure Mathematics)University of Waterloo, Canada 

  • M.Sc., Computer ScienceUniversity of Toronto, Canada 

  • Ph.D., Computer ScienceUniversity of Toronto (Advisor: Geoffrey Hinton), Canada 



His Ph.D. thesis was “Bethe free energy and contrastive divergence approximations for undirected graphical models”. 



3. Academic & Professional Experience


  • Professor of Statistical Machine Learning, University of Oxford (present) 

  • Principal Research Scientist / Research Director, DeepMind (present) 

  • Lecturer & Reader, Gatsby Computational Neuroscience Unit, University College London (2007–2012) 

  • Postdoctoral Researcher, University of California, Berkeley and National University of Singapore (Lee Kuan Yew Fellow) 




4. Research Interests



Yee Whye Teh’s work focuses on:

  • Probabilistic machine learning

  • Bayesian nonparametrics & hierarchical models

  • Variational inference & Monte Carlo methods

  • Deep learning theory and scalable inference

  • Applications in AI, genetics, linguistics, neuroscience & statistics 




5. Honors & Service


  • IMS Medallion Lecture, Joint Statistical Meetings 2019 

  • Breiman Lecture, NeurIPS 2017 

  • Program Co-Chair — International Conference on Machine Learning (ICML) 2017, AISTATS 2010 

  • Editorial & area chair roles at IEEE TPAMI, JMLR, Bayesian Analysis, JRSS Series B, etc. 

  • ELLIS Fellow and co-director roles promoting robust machine learning across Europe. 




6. Notable Contributions



Teh is one of the original developers of deep belief networks and hierarchical Dirichlet processes, and has influenced both theoretical and applied aspects of machine learning and statistical AI