多維時(shí)序 | Matlab實(shí)現(xiàn)LSTM-Adaboost和LSTM多變量時(shí)間序列預(yù)測(cè)對(duì)比

預(yù)測(cè)效果

基本介紹

多維時(shí)序 | Matlab實(shí)現(xiàn)LSTM-Adaboost和LSTM多變量時(shí)間序列預(yù)測(cè)對(duì)比

模型描述

Matlab實(shí)現(xiàn)LSTM-Adaboost和LSTM多變量時(shí)間序列預(yù)測(cè)對(duì)比（完整程序和數(shù)據(jù)）
1.輸入多個(gè)特征，輸出單個(gè)變量；
2.考慮歷史特征的影響，多變量時(shí)間序列預(yù)測(cè)；
4.csv數(shù)據(jù)，方便替換；
5.運(yùn)行環(huán)境Matlab2018b及以上；
6.輸出誤差對(duì)比圖。

程序設(shè)計(jì)

• 完整程序和數(shù)據(jù)獲取方式1：同等價(jià)值程序兌換；
• 完整程序和數(shù)據(jù)獲取方式2：私信博主回復(fù)Matlab實(shí)現(xiàn)LSTM-Adaboost和LSTM多變量時(shí)間序列預(yù)測(cè)對(duì)比獲取
• 完整程序和數(shù)據(jù)獲取方式3（直接下載）：Matlab實(shí)現(xiàn)LSTM-Adaboost和LSTM多變量時(shí)間序列預(yù)測(cè)對(duì)比。

 (32,'OutputMode',"last",'Name','bil4','RecurrentWeightsInitializer','He','InputWeightsInitializer','He')dropoutLayer(0.25,'Name','drop2')% 全連接層fullyConnectedLayer(numResponses,'Name','fc')regressionLayer('Name','output')    ];layers = layerGraph(layers);layers = connectLayers(layers,'fold/miniBatchSize','unfold/miniBatchSize');
%-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
%% 訓(xùn)練選項(xiàng)
if gpuDeviceCount>0mydevice = 'gpu';
elsemydevice = 'cpu';
endoptions = trainingOptions('adam', ...'MaxEpochs',MaxEpochs, ...'MiniBatchSize',MiniBatchSize, ...'GradientThreshold',1, ...'InitialLearnRate',learningrate, ...'LearnRateSchedule','piecewise', ...'LearnRateDropPeriod',56, ...'LearnRateDropFactor',0.25, ...'L2Regularization',1e-3,...'GradientDecayFactor',0.95,...'Verbose',false, ...'Shuffle',"every-epoch",...'ExecutionEnvironment',mydevice,...'Plots','training-progress');
%% 模型訓(xùn)練
rng(0);
net = trainNetwork(XrTrain,YrTrain,layers,options);
%-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
%% 測(cè)試數(shù)據(jù)預(yù)測(cè)
% 測(cè)試集預(yù)測(cè)
YPred = predict(net,XrTest,"ExecutionEnvironment",mydevice,"MiniBatchSize",numFeatures);
YPred = YPred';
% 數(shù)據(jù)反歸一化
YPred = sig.*YPred + mu;
YTest = sig.*YTest + mu;
————————————————
版權(quán)聲明：本文為CSDN博主「機(jī)器學(xué)習(xí)之心」的原創(chuàng)文章，遵循CC 4.0 BY-SA版權(quán)協(xié)議，轉(zhuǎn)載請(qǐng)附上原文出處鏈接及本聲明。

參考資料

[1] http://t.csdn.cn/pCWSp
[2] https://download.csdn.net/download/kjm13182345320/87568090?spm=1001.2014.3001.5501
[3] https://blog.csdn.net/kjm13182345320/article/details/129433463?spm=1001.2014.3001.5501