Abstract Languages vary in how they structure the terms for a semantic domain, such as colors or spatial relations. For example, in English we say “the cup is on the table”, “the ring is on[…]
Abstract Performance of machine learning systems depends critically on tuning parameters that are difficult to set by standard optimization techniques. Such “hyperparamers”—including model architecture, regularization, and learning rates—are often tuned in an outerloop by black-box[…]
Abstract We present Sparse Non-negative Matrix (SNM), a novel probability estimation technique for language modeling that can efficiently incorporate arbitrary features in a similar way to the more established family of maximum entropy (exponential models). […]
Abstract (Joint work with Shlomo E. Chazan and Jacob Goldberger) In this work, we propose a hybrid approach for single microphone speech enhancement, merging the generative Mixture of Gaussians (MoG) model and the discriminative deep[…]
Abstract In this talk, I start with a brief introduction to the history of symbolic approaches to natural language processing (NLP), and why we move to neural approaches recently. Then I describes in detail the[…]
Abstract In this talk, I’ll discuss two real-world language applications that require “thinking on your feet”: synchronous machine translation (or “machine simultaneous interpretation”) and question answering (when questions are revealed one piece at a time). […]
Abstract Recent advances in deep learning based approaches enabled exciting new research frontiers for goal-oriented conversation systems. In this talk I will present an end-to-end dialog system, with components for language understanding, dialogue state tracking, policy,[…]
