matlab用Lstm空气质量预测代码

以下是一个基于 LSTM 的空气质量预测的 MATLAB 代码示例：

% 读取数据
data = csvread('air_quality.csv',1,0);
dates = datetime(data(:,1:3));
x = datenum(dates);
y = data(:,4);

% 分割数据集
train_ratio = 0.8;
train_num = floor(train_ratio * length(x));
x_train = x(1:train_num);
y_train = y(1:train_num);
x_test = x(train_num+1:end);
y_test = y(train_num+1:end);

% 构建 LSTM 模型
num_features = 1;
num_hidden_units = 100;
dropout_rate = 0.2;
num_epochs = 50;
mini_batch_size = 64;
layers = [ ...
    sequenceInputLayer(num_features)
    lstmLayer(num_hidden_units,'OutputMode','last')
    dropoutLayer(dropout_rate)
    fullyConnectedLayer(1)
    regressionLayer];
options = trainingOptions('adam', ...
    'MaxEpochs',num_epochs, ...
    'MiniBatchSize',mini_batch_size, ...
    'GradientThreshold',1, ...
    'Shuffle','never', ...
    'Verbose',0);
net = trainNetwork(x_train',y_train',layers,options);

% 预测测试集
y_pred = predict(net,x_test')';

% 计算均方误差
mse = mean((y_pred - y_test).^2);
fprintf('均方误差：%f\n',mse);

% 绘制预测结果
figure;
plot(x,y,'b.-');
hold on;
plot(x_test,y_pred,'r.-');
legend('实际值','预测值');
grid on;
datetick('x','yyyy-mm-dd','keepticks');
xlabel('日期');
ylabel('AQI');

其中，`air_quality.csv` 是一个包含日期和 AQI 值的 CSV 文件。此代码将数据集分为训练集和测试集，使用 LSTM 模型进行训练和预测，并计算均方误差。最后，绘制实际值和预测值的折线图。