期刊
NEURAL COMPUTATION
卷 20, 期 10, 页码 2526-2563出版社
MIT PRESS
DOI: 10.1162/neco.2008.03-07-486
关键词
-
资金
- NSF [IIS-06-25223, CCF-0431150]
- DARPA/ONR [N66001-06-1-2011, N00014-06-1-0610]
- ONR [N00014-06-1-0769, N00014-06-1-0829]
- AFOSR [FA9550-04-1-0148]
- Texas Instruments DSP Leadership University Program
- [NGAMCA015894-UCB]
While evidence indicates that neural systems may be employing sparse approximations to represent sensed stimuli, the mechanisms underlying this ability are not understood. We describe a locally competitive algorithm (LCA) that solves a collection of sparse coding principles minimizing a weighted combination of mean-squared error and a coefficient cost function. LCAs are designed to be implemented in a dynamical system composed of many neuron-like elements operating in parallel. These algorithms use thresholding functions to induce local (usually one-way) inhibitory competitions between nodes to produce sparse representations. LCAs produce coefficients with sparsity levels comparable to the most popular centralized sparse coding algorithms while being readily suited for neural implementation. Additionally, LCA coefficients for video sequences demonstrate inertial properties that are both qualitatively and quantitatively more regular (i.e., smoother and more predictable) than the coefficients produced by greedy algorithms.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据