L
Lawrence K. Saul
Researcher at University of California, San Diego
Publications - 138
Citations - 40154
Lawrence K. Saul is an academic researcher from University of California, San Diego. The author has contributed to research in topics: Hidden Markov model & Nonlinear dimensionality reduction. The author has an hindex of 49, co-authored 133 publications receiving 37255 citations. Previous affiliations of Lawrence K. Saul include Massachusetts Institute of Technology & University of Pennsylvania.
Papers
More filters
Journal ArticleDOI
Attractor Dynamics in Feedforward Neural Networks
TL;DR: This work establishes global convergence of the dynamics by providing a Lyapunov function and shows that the dynamics generate the signals required for unsupervised learning.
Patent
Speech and speaker recognition using factor analysis to model covariance structure of mixture components
Mazin G. Rahim,Lawrence K. Saul +1 more
TL;DR: In this article, factor analysis is used to model acoustic correlation in automatic speech recognition by introducing a small number of parameters to model the covariance structure of a speech signal, which are estimated by an Expectation Maximization (EM) technique that can be embedded in the training procedures for the HMMs, and then further adjusted using Minimum Classification Error (MCE) training, which demonstrates better discrimination and produces more accurate recognition models.
Proceedings Article
Exploiting Feature Covariance in High-Dimensional Online Learning
TL;DR: This work develops low-rank approximations of the covariance structure of the weights of second-order algorithms for linear classification and shows improvements over diagonal covariance matrices for both low and high-dimensional data.
Journal ArticleDOI
Robust numeric recognition in spoken language dialogue
Mazin G. Rahim,Giuseppe Riccardi,Lawrence K. Saul,Jerry H. Wright,Bruce Buntschuh,Allen Louis Gorin +5 more
TL;DR: A robust system for numeric recognition is described and algorithms for feature extraction, acoustic and language modeling, discriminative training, utterance verification and numeric understanding and validation are presented.
Proceedings ArticleDOI
Generating correctness proofs with neural networks
TL;DR: Proverbot9001 as mentioned in this paper is a proof search system using machine learning techniques to produce proofs of software correctness in interactive theorem provers, which can effectively automate what were previously manual proofs, automatically producing proofs for 28% of theorem statements in our test dataset, when combined with solver-based tooling.