Time series prediction | MATLAB implements BO-BiGRU Bayesian optimization bidirectional gated cycle unit time series prediction

Time series prediction | MATLAB implements BO-BiGRU Bayesian optimization bidirectional gated cycle unit time series prediction

Effect list

basic introduction

MATLAB implements BO-BiGRU Bayesian optimization bidirectional gated cycle unit time series prediction. Time series prediction based on Bayesian optimization of bidirectional gated cyclic units, BO-BiGRU/Bayes-BiGRU time series prediction model.
1. The optimization parameters are: learning rate, hidden layer nodes, and regularization parameters.
2. Evaluation indicators include: R2, MAE, MSE, RMSE and MAPE, etc.
3. The operating environment is matlab2018b and above.

Model building

BO-BiGRU (Bayesian Optimization Bidirectional Gated Recurrent Unit) is a method that combines Bayesian optimization and Bidirectional Gated Recurrent Unit (BiGRU) for time series forecasting tasks. In time series forecasting, we try to predict future values ​​based on past observations.
Bidirectional gated recurrent unit (BiGRU) is a variant of recurrent neural network (RNN) with more powerful modeling capabilities than traditional recurrent neural networks. It better captures long-term dependencies in time series by using update gates and reset gates to control the flow of information.
Bayesian optimization is a method for optimization problems that can sample on an unknown objective function and adjust the sampling position based on existing samples. This method can help us find the optimal solution efficiently in the search space.
The basic idea of ​​BO-BiGRU is to use Bayesian optimization to automatically adjust the hyperparameters of the model to obtain better time series prediction performance. The Bayesian optimization algorithm selects the next hyperparameter configuration for evaluation based on existing model performance samples, gradually searches the hyperparameter space, and uses the Bayesian inference method to update the probability distribution of the hyperparameters. In this way, BO-BiGRU can find better hyperparameter configurations within a relatively small number of model training iterations, thereby improving the accuracy of time series predictions.
To summarize, BO-BiGRU is a method that combines Bayesian optimization and gated recurrent units for time series forecasting tasks. It improves model performance by automatically tuning hyperparameters and is able to better capture long-term dependencies in time series.

• MATLAB implements BO-BiGRU Bayesian optimization bidirectional gated cycle unit time series prediction
  pseudo code
  
• Adjust model parameters by adjusting optimization algorithms, learning repetition rates, and Bayesian optimization hyperparameters to adjust model parameters.

programming

• Complete program and data acquisition method: Private message the blogger to reply to MATLAB to implement BO-GRU Bayesian optimization gated cycle unit time series prediction .

%%  优化算法参数设置
%参数取值上界(学习率，隐藏层节点，正则化系数)
%%  贝叶斯优化参数范围
optimVars = [
    optimizableVariable('NumOfUnits', [10, 50], 'Type', 'integer')
    optimizableVariable('InitialLearnRate', [1e-3, 1], 'Transform', 'log')
    optimizableVariable('L2Regularization', [1e-10, 1e-2], 'Transform', 'log')];

%% 创建网络架构
% 输入特征维度
numFeatures  = f_;
% 输出特征维度
numResponses = 1;
FiltZise = 10;
%  创建"LSTM"模型
    layers = [...
        % 输入特征
        sequenceInputLayer([numFeatures 1 1],'Name','input')
        sequenceFoldingLayer('Name','fold')
        % 特征学习       

        dropoutLayer(0.25,'Name','drop3')
        % 全连接层
        fullyConnectedLayer(numResponses,'Name','fc')
        regressionLayer('Name','output')    ];

    layers = layerGraph(layers);
    layers = connectLayers(layers,'fold/miniBatchSize','unfold/miniBatchSize');


% 批处理样本
MiniBatchSize =128;
% 最大迭代次数
MaxEpochs = 500;
    options = trainingOptions( 'adam', ...
        'MaxEpochs',500, ...
        'GradientThreshold',1, ...
        'InitialLearnRate',optVars.InitialLearnRate, ...
        'LearnRateSchedule','piecewise', ...
        'LearnRateDropPeriod',400, ...
        'LearnRateDropFactor',0.2, ...
        'L2Regularization',optVars.L2Regularization,...
        'Verbose',false, ...
        'Plots','none');

%% 训练混合网络
net = trainNetwork(XrTrain,YrTrain,layers,options);

References

[1] https://blog.csdn.net/kjm13182345320/article/details/129036772?spm=1001.2014.3001.5502
[2] https://blog.csdn.net/kjm13182345320/article/details/128690229