提高LSTM模型的准确性

Question

嗨，我正在尝试使用LSTM模型训练数据。我的数据集具有2个特征和16个类。

我有几个问题： 1：训练和验证中的准确性都是波动。我应该如何解决这个问题。 2：验证数据的准确性没有提高

 numFeatures = 2;
    numHiddenUnits = 250;
    numClasses = 16;

    options = trainingOptions('adam', ...
    'ExecutionEnvironment','gpu', ...
    'InitialLearnRate',0.005, ...
    'LearnRateSchedule','piecewise', ...
    'GradientThreshold',2, ...
    'SequenceLength',10000,...
    'MaxEpochs',maxEpochs, ...
    'MiniBatchSize',miniBatchSize, ...
    'Shuffle','never', ...
    'ValidationData',{Input_Val,Output_Val}, ...
    'ValidationFrequency',5, ...
    'Verbose',0, ...
    'Plots','training-progress');

    layers = [ ...
    sequenceInputLayer(numFeatures)
    lstmLayer(numHiddenUnits,'OutputMode','sequence')
    dropoutLayer(0.2)
    lstmLayer(numHiddenUnits,'OutputMode','sequence')
    dropoutLayer(0.2)
   lstmLayer(numHiddenUnits,'OutputMode','sequence')
    dropoutLayer(0.2)
    fullyConnectedLayer(numClasses)
    softmaxLayer
    classificationLayer];

以下是准确性和误差图：