Taking Humans out of the Deep Learning Loop
Machine learning frequently involves careful tuning of learning parameters and model hyperparameters. Unfortunately, this tuning is often a "black art" requiring expert experience and heuristics. This pain is particularly acute in deep learning and there is great appeal in automatic approaches that optimize learning algorithms for the problem at hand. I will discuss how this can be tackled with Bayesian optimization, surpassing human experts on many deep learning problems. I'll also talk about applying these ideas across the sciences, and discuss how our new startup, Whetlab, is making it easy to use Bayesian optimization over the wire.
Ryan Adams is an Assistant Professor of Computer Science at Harvard. He received his Ph.D. in Physics at Cambridge as a Gates Scholar. He was a CIFAR Junior Research Fellow at the University of Toronto before joining the faculty at Harvard. His Ph.D. thesis received Honorable Mention for the Savage Award for Theory and Methods from the International Society for Bayesian Analysis. Dr. Adams has won paper awards at the International Conference on Machine Learning, the International Conference on Artificial Intelligence and Statistics, and the Conference on Uncertainty in Artificial Intelligence. He has also received the DARPA Young Faculty Award and the Sloan Fellowship. Dr. Adams is the CEO of Whetlab, a machine learning startup, and co-hosts the popular Talking Machines podcast.