预测序列的特征维度1但输入层预计序列的特征维度2。
14日视图(30天)
显示旧的评论
大家好
我试图预测交通流量通过使用实时数据和以前的数据。所以我有两个输入和一个输出。我使用LSTM为此,然后当我想预测数据使用“YPred =预测(净、欧美);“但我面对一个错误预测序列的特征维度1但输入层预计序列的特征维度2。
我XTrain 470列({2×1双})和YTrain有479列1 * 1双。
XTest和欧美有60列细胞是一样的火车
我尝试一切但我不能修好它
请帮我
谢谢
我的代码是:
clc
关闭所有
清晰的所有
警告从
% [~,~,flow_data] = xlsread (“two_days.xlsx”);
flow_data = xlsread (“two_days.xlsx”);%这里我们有两天的数据
% data_mat = cell2mat (flow_data(2:最终,3:4));
data_mat = (flow_data(2:最终,3:4));
% XTrain = data_mat (:, 3:4) ';
% YTrain = data_mat (:, 3:4) ';
XTrain = data_mat (:,:)”;
YTrain = data_mat (: 1)”;
XTrain = num2cell (XTrain, 1);
YTrain = num2cell (YTrain, 1);
% %
numResponses = 1;
% numResponses =大小(YTrain {1}, 1);
featureDimension = 2;
numHiddenUnits = 200;
层= […
sequenceInputLayer (featureDimension)
lstmLayer (numHiddenUnits)
% dropoutLayer (0.1) % % 0.5
fullyConnectedLayer (numResponses)
regressionLayer];
maxepochs = 500;
miniBatchSize = 1;
选择= trainingOptions (“亚当”,…% %亚当
“MaxEpochs”maxepochs,…
“GradientThreshold”,1…
“InitialLearnRate”,0.005,…
“LearnRateSchedule”,“分段”,…
“LearnRateDropPeriod”,225,…
“LearnRateDropFactor”,0.2,…
“详细”,1…
“阴谋”,“训练进步”);
% % %训练网络
网= trainNetwork (XTrain、YTrain层,选择);
%
test_data = xlsread (“test_data2.xlsx”);%这里我们有两天的数据
data_mat2 = (test_data(1:最终,3:4));
% XTest = data_mat2 (:,:)”;
欧美= data_mat2 (: 1)”;
XTest = data_mat2 (: 1)”;
XTrain = num2cell (XTest, 1);
YTrain = num2cell(1次);
网= resetState(净);
YPred =预测(净、欧美);
YPred =圆(YPred)
% =净predictAndUpdateState(网络,XTrain);
%(网络,YPred) = predictAndUpdateState(净,YTrain(结束));
% %
% % %
% %预测只要测试周期
% numTimeStepsTest =元素个数(欧美);
% i = 2: numTimeStepsTest
%(净,YPred(:,我)]= predictAndUpdateState(净,YPred(张:,),“ExecutionEnvironment”,“cpu”);
%结束
% YPred
%日元= (cell2mat (YPred(1: 1:末端)));%有阴谋情节列转置
%的阴谋(日元)
%抓住
% y2 = (cell2mat(欧美(1:1:末端))");
%的阴谋(y2)
