% 1. 数据准备
X_train = 训练数据输入;
Y_train = 训练数据输出;
X_test = 测试数据输入;

% 2. 模型构建
inputSize = size(X_train, 2);
numHiddenUnits = 100;
numResponses = 1;

layers = [ …
sequenceInputLayer(inputSize)
biLSTMLayer(numHiddenUnits, ‘OutputMode’, ‘sequence’)
dropoutLayer(0.2)
fullyConnectedLayer(numResponses)
regressionLayer
];

options = trainingOptions(‘adam’, …
‘MaxEpochs’,50, …
‘MiniBatchSize’, 32, …
‘GradientThreshold’, 1, …
‘SequenceLength’, 20, …
‘Plots’,‘training-progress’);

% 3. 贝叶斯优化
vars = [
optimizableVariable(‘MiniBatchSize’,[32, 128],‘Type’,‘integer’)
optimizableVariable(‘SequenceLength’,[10, 30],‘Type’,‘integer’)
];

ObjFcn = @(params)trainBiGRU(params, X_train, Y_train, layers, options);
results = bayesopt(ObjFcn, vars, ‘MaxObjectiveEvaluations’, 30);

% 4. 训练模型
bestParams = bestPoint(results);
bestMiniBatchSize = bestParams.MiniBatchSize;
bestSequenceLength = bestParams.SequenceLength;

options.MiniBatchSize = bestMiniBatchSize;
options.SequenceLength = bestSequenceLength;

net = trainNetwork(X_train, Y_train, layers, options);

% 5. 模型评估
YPred = predict(net, X_test);

% 6. 预测
disp(YPred);
————————————————

                        版权声明：本文为博主原创文章，遵循 CC 4.0 BY-SA 版权协议，转载请附上原文出处链接和本声明。

原文链接：https://blog.csdn.net/qq_59771180/article/details/143499678