conftrace_

2012 NeurIPS NeurIPS 2012

Proximal Newton-type methods for convex optimization

Abstract

We seek to solve convex optimization problems in composite form:

🌉 Interdisciplinary Bridge - Artificial Intelligence and Mathematics & Optimization

📈 Trend Setter - Optimization

🧭 Keyword Pioneer - first-order method

Authors

Jason Lee, Yuekai Sun, Michael Saunders

Topics

Machine Learning > Optimization & Theory > Optimization Mathematics & Optimization > Optimization > Continuous Optimization Deep Learning > Optimization & Theory > Optimization Mathematics & Optimization > Optimization > Convex Optimization

Keywords

convex optimization composite optimization proximal newton methods first-order method newton-type method newton method proximal method second order cone program

More from NeurIPS 2012

All NeurIPS 2012 papers →

Efficient high dimensional maximum entropy modeling via symmetric partition functions

Topic-Partitioned Multinetwork Embeddings

Recovery of Sparse Probability Measures via Convex Programming

Learning visual motion in recurrent neural networks

Graphical Models via Generalized Linear Models

Searching for objects driven by context