ample-master

所属分类:人工智能/神经网络/深度学习
开发工具:Python
文件大小:245KB
下载次数:3
上传日期:2020-11-05 19:55:47
上 传 者c罗145
说明:  解决回归问题,利用近似消息传递解决回归问题。。。。。。。。。。。。。
(Solving the problem of return)

文件列表:
LICENSE (18028, 2015-04-02)
ample.m (14414, 2015-04-02)
demos (0, 2015-04-02)
demos\ample_binary_demo.m (3686, 2015-04-02)
demos\ample_em_demo.m (4090, 2015-04-02)
demos\ample_em_fourier_demo.m (4009, 2015-04-02)
demos\ample_fourier_demo.m (4851, 2015-04-02)
demos\ample_gb_demo.m (5052, 2015-04-02)
demos\ample_qary_demo.m (4454, 2015-04-02)
demos\random_scrambled_fourier.m (2154, 2015-04-02)
doc (0, 2015-04-02)
doc\ample_manual.html (24778, 2015-04-02)
doc\ample_manual.md (13231, 2015-04-02)
doc\ample_manual.pdf (213324, 2015-04-02)
priors (0, 2015-04-02)
priors\prior_binary.m (1065, 2015-04-02)
priors\prior_gb.m (1781, 2015-04-02)
priors\prior_l1sparse.m (1577, 2015-04-02)
priors\prior_qary.m (2893, 2015-04-02)

ample ===== A General implementation of Approximate Message Passing (AMP) for Matlab. ***Note:*** This repository is still under active development. There may be some hiccups. Intro ----- The goal of this problem is to provide a quick-to-use and easily generalizable implementation of the approximate message passing (AMP) algorithm.... * ***Quick-to-use:*** At its core, one only needs to provide the set of obstervations, the system which obtained those observations, and the type of signal prior. However, `ample` contains a number of options to allow the user to tailor AMP to suit their particular needs. * ***Generalizable:*** `ample` provides a framework by which the user can easily implement their own priors to fit AMP to their specific application. Acknowledgements ---------------- Many thanks to Jean Barbier and Andre Manoel for their moment and update calculations for the supplied priors. Also, special thanks to Florent Krzakala, whose ERC SPARCS 307087 grant funded this work.

近期下载者

相关文件


收藏者