Scholar
James Martens
Google Scholar ID: LlK_saMAAAAJ
Research Scientist, DeepMind
Citations & Impact
Citations
15,142
H-index
29
i10-index
33
Publications
20
Co-authors
0
Contact
Publications
Resume
Academic Achievements
- K-FAC, a powerful new optimization method for neural networks.
Research Experience
- Joining Google-DeepMind in September 2016.
Education
- Ph.D. in Computer Science from the University of Toronto, supervised by Geoffrey Hinton and Richard Zemel.
Background
- Research Interests: Development of learning algorithms, particularly the application of 2nd-order optimization techniques to deep and otherwise hard-to-optimize models; sampling and inference algorithms; expressive efficiency of deep models; unsupervised learning.
Miscellany
- CV last updated: Feb 2016.