Matlab中实现了利用RNN-LSTM的卷积神经网络

function [train_datatest_data]=LSTM_data_process（）
%% 数据加载并完成初始归一化
train_data_initial= [0.4413 0.4707 0.6953 0.8133 0.4379 0.4677 0.6981 0.8002 0.4517 0.4725 0.7006 0.8201;
                     0.4379 0.4677 0.6981 0.8002 0.4517 0.4725 0.7006 0.8201 0.4557 0.4790 0.7019 0.8211;
                     0.4517 0.4725 0.7006 0.8201 0.4557 0.4790 0.7019 0.8211 0.4601 0.4911 0.7101 0.8298]‘;
% train_data_initial=[ 0.4413 0.4707 0.6953 0.8133;
%                      0.4379 0.4677 0.6981 0.8002;
%                      0.4517 0.4725 0.7006 0.8201;
%                      0.4557 0.4790 0.7019 0.8211;
%                      0.4601 0.4811 0.7101 0.8298;
%                      0.4612 0.4845 0.7188 0.8312]‘;
test_data_initial=[0.4557 0.4790 0.7019 0.8211;
                   0.4612 0.4845 0.7188 0.8312;
                   0.4601 0.4811 0.7101 0.8298;
                   0.4615 0.4891 0.7201 0.8330]‘;

data_length=size（train_data_initial1）;            %每个样本的长度
data_num=size（train_data_initial2）;               %样本数目  

%%归一化过程
for n=1:data_num
    train_data（:n）=train_data_initial（:n）/sqrt（sum（train_data_initial（:n）.^2））;  
end
for m=1:size（test_data_initial2）
    test_data（:m）=test_data_initial（:m）/sqrt（sum（test_data_initial（:m）.^2））;
end

 属性            大小     日期    时间   名称
----------- ---------  ---------- -----  ----
     目录           0  2018-11-30 16:53  LSTM_MY_updata_to_precise\
     文件        1301  2016-10-08 12:12  LSTM_MY_updata_to_precise\LSTM_data_process.m
     文件        6071  2016-10-12 15:18  LSTM_MY_updata_to_precise\LSTM_mian.m
     文件        4544  2016-10-11 16:01  LSTM_MY_updata_to_precise\LSTM_updata_weight.m