A
Andrew Cotter
Researcher at Google
Publications - 45
Citations - 3847
Andrew Cotter is an academic researcher from Google. The author has contributed to research in topics: Constrained optimization & Support vector machine. The author has an hindex of 22, co-authored 45 publications receiving 3436 citations. Previous affiliations of Andrew Cotter include Toyota Technological Institute at Chicago.
Papers
More filters
Journal ArticleDOI
Pegasos: primal estimated sub-gradient solver for SVM
TL;DR: A simple and effective stochastic sub-gradient descent algorithm for solving the optimization problem cast by Support Vector Machines, which is particularly well suited for large text classification problems, and demonstrates an order-of-magnitude speedup over previous SVM learning methods.
Proceedings Article
Better Mini-Batch Algorithms via Accelerated Gradient Methods
TL;DR: A novel analysis is provided, which shows how standard gradient methods may sometimes be insufficient to obtain a significant speed-up and a novel accelerated gradient algorithm is proposed, which deals with this deficiency, enjoys a uniformly superior guarantee and works well in practice.
Proceedings ArticleDOI
Stochastic optimization for PCA and PLS
TL;DR: Several stochastic approximation methods for PCA and PLS are suggested, and empirical performance of these methods is investigated.
Proceedings Article
Satisfying real-world goals with dataset constraints
TL;DR: In this article, the authors propose handling multiple goals on multiple datasets by training with dataset constraints, using the ramp penalty to accurately quantify costs, and present an efficient algorithm to approximately optimize the resulting non-convex constrained optimization problem.