DNN
所属分类：数值算法/人工智能
开发工具：matlab
文件大小：41077KB
下载次数：24
上传日期：2018-04-13 10:21:22
上 传 者：
漫长丹霞
说明： 基于WMMSE的DNN算法，通过WMMSE获取动态的学习率以自适应网络
(this is a simple code collection for DNN which based on WMMSE)
文件列表:
DNN (0, 2018-04-09)
DNN\DNN_CDF_10_50000_20_20.fig (1213304, 2018-04-09)
DNN\drawfig.m (1051, 2017-10-14)
DNN\generate.m (1164, 2017-10-14)
DNN\Gussianfit_10_50000_20_20.m (57565, 2018-04-09)
DNN\Gussian_50000_10.mat (40817002, 2018-04-09)
DNN\main.m (1233, 2017-10-14)
DNN\obj_IA_sum_rate.m (871, 2017-10-14)
DNN\testperformance.m (2261, 2017-10-14)
DNN\trainDNN.m (1881, 2017-10-14)
DNN\trainperformance.m (1402, 2017-10-14)
DNN\WMMSE_sum_rate.m (1456, 2017-10-14)
# DNN_WMMSE
[Update]: This code is outdated, please refer to our Python version: https://github.com/Haoran-S/SPAWC2017.
---------------------------------------------------------------------
MATLAB code to reproduce our works on DNN research.
Simply run "main.m", you will get the result for Gaussian IC case.
To get results for other sections, slightly modification may apply.
We also provide some pre-trained functions to show our results in Table. 1 & Table 2.
To run our code, Neuron Network Toolbox and Deep Learning Toolbox need to be installed first.
Code has been tested successfully on MATLAB 2016b prerelease platform.
References:
[1] Haoran Sun, Xiangyi Chen, Qingjiang Shi, Mingyi Hong, Xiao Fu, Nikos D. Sidiropoulos.
"LEARNING TO OPTIMIZE: TRAINING DEEP NEURAL NETWORKS FOR WIRELESS RESOURCE MANAGEMENT."
version 1.0 -- September/2016
Written by Haoran Sun (hrsun AT iastate.edu)
近期下载者：
相关文件：
收藏者：