We present a novel approach to dealing with overfitting in black-box models. It is based on the leverages of the samples, i.e. on the influence that each observation has on the pa...
This paper presents an algorithm for an 1-regularized Kalman filter. Given observations of a discrete-time linear dynamical system with sparse errors in the state evolution, we e...
Muhammad Salman Asif, Adam Charles, Justin K. Romb...
Wireless technologies are poised to bring major changes in the information infrastructure, transforming communications and facilitating information retrieval. Over the next severa...
For many supervised learning tasks it is very costly to produce training data with class labels. Active learning acquires data incrementally, at each stage using the model learned...
We present BL-WoLF, a framework for learnability in repeated zero-sum games where the cost of learning is measured by the losses the learning agent accrues (rather than the number...