Portrait of Ofer Dekel

Ofer Dekel

Partner Research Manager

About

I manage the Machine Learning and Optimization group (opens in new tab), at Microsoft Research, in Redmond, Washington. My main research areas are machine learning algorithms, online prediction, algorithm engineering, statistical learning theory, and optimization.

In the last couple of years, I’ve been focusing on machine learning algorithms and technologies tailored for tiny resource-constrained computers (opens in new tab), like the ones embedded in intelligent devices. I’m interested in algorithms that compress large existing models, such as deep convolutional neural networks. I’m also interested in finding new prediction models (not necessarily deep neural networks) that are specifically tailored for resource impoverished platforms.

My team designs and builds the Embedded Learning Library (ELL) (opens in new tab), which is a set of tools that allows developers to design and deploy intelligent machine-learned models onto resource constrained platforms and small single-board computers, like Raspberry Pi.

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.

My Publications (opens in new tab)