Sparse Online Learning via Truncated Gradient

13 years 5 months ago

Download jmlr.csail.mit.edu

We propose a general method called truncated gradient to induce sparsity in the weights of onlinelearning algorithms with convex loss functions. This method has several essential properties:

John Langford, Lihong Li, Tong Zhang

Real-time Traffic

Convex Loss Functions | Essential Properties | General Method | Information Technology | NIPS 2008 |

claim paper

» Learning incoherent sparse and lowrank patterns from multiple tasks

» Denoising sparse noise via online dictionary learning

» NonlinearDynamical Attention Allocation via Information Geometry

» Online Detection of Unusual Events in Videos via Dynamic Sparse Coding

» Adaptive Subgradient Methods for Online Learning and Stochastic Optimization

» Sparse incremental learning for interactive robot control policy estimation

» Proximal Methods for Sparse Hierarchical Dictionary Learning

» Dual Averaging Methods for Regularized Stochastic Learning and Online Optimization

Post Info
More Details (n/a)

Added	29 Oct 2010
Updated	29 Oct 2010
Type	Conference
Year	2008
Where	NIPS
Authors	John Langford, Lihong Li, Tong Zhang

Comments (0)

Sciweavers

Sparse Online Learning via Truncated Gradient

Convex Loss Functions | Essential Properties | General Method | Information Technology | NIPS 2008 |

Explore & Download

Productivity Tools

Document Tools

Image Tools

Sciweavers