Optimization methods are the engine of machine learning algorithms. Examples abound, such as training neural networks with stochastic gradient descent, segmenting images with submodular optimization, or efficiently searching a game tree with bandit algorithms. We aim to advance the mathematical foundations of both discrete and continuous optimization and to leverage these advances to develop new algorithms with a broad set of AI applications. Some of the current directions pursued by our members include convex optimization,…
I manage the Machine Learning and Optimization group, at Microsoft Research, in Redmond, Washington. My main research interests are machine learning, online prediction, algorithm engineering, statistical learning theory, and optimization.
Recently, I’ve been focusing on machine learning algorithms and technologies tailored for tiny resource-constrained computers, like the ones embedded in intelligent devices. I’m interested in algorithms that compress large existing models, such as deep neural networks. I’m also interested in new prediction models that are particularly designed for resource impoverished platforms.
In the past, I have worked on online learning and bandit algorithms, support vector machines, boosting algorithms, online to batch conversion techniques, incentive compatible learning, learning from multiple teachers and crowdsourced data, growing and pruning decision trees, extreme classification, and other topics in learning and optimization.
In a few years, the world will be filled with billions of small, connected, intelligent devices. Many of these devices will be embedded in our homes, our cities, our vehicles, and our factories. Some of these devices will be carried in our pockets or worn on our bodies. The proliferation of small computing devices will disrupt every industrial sector and play a key role in the next evolution of personal computing. Most of these devices…
March 27, 2015
Machine Learning Work Shop – Session 2 – Ofer Dekel – “Online Learning Against Adaptive Adversaries”
October 26, 2012
MSR - Machine Learning Group