Artificial Intelligence
Pursuing computing advances to create intelligent machines that complement human reasoning to augment and enrich our experience and competencies.
Project
Filter Forests for Learning Data-Dependent Convolutional Kernels
We propose ‘filter forests’ (FF), an efficient new discriminative approach for predicting continuous variables given a signal and its context. FF can be used for general signal restoration tasks that can be tackled via convolutional…
Publication
Adapting Deep RankNet for Personalized Search
Project
Spatial Crowdsourcing
We are studying how we can get regular people to do simple tasks at specific locations. An example task is to take a picture of a sign at a certain location. We are interested in…
Project
Learning Theory
We work on questions motivated by machine learning, in particular from the theoretical and computational perspectives. Our goals are to mathematically understand the effectiveness of existing learning algorithms and to design new learning algorithms. We…