LSTM分类双向长短时记忆BiLSTM实现数据分类含Matlab源码

%加载序列数据
%数据描述：总共270组训练样本共分为9类，每组训练样本的训练样个数不等，每个训练训练样本由12个特征向量组成，
clc
clear all
close all
[XTrain,YTrain] = japaneseVowelsTrainData;
%数据可视化
figure
plot(XTrain{1}')
xlabel('Time Step')
title('Training Observation 1')
legend('Feature ' ,'Location','northeastoutside')
%%
%LSTM可以将分组后等量的训练样本进行训练，从而提高训练效率
%如果每组的样本数量不同，进行小批量拆分，则需要尽量保证分块的训练样本数相同
%首先找到每组样本数和总的组数
numObservations = numel(XTrain);
for i=1:numObservations
    sequence = XTrain{i};
    sequenceLengths(i) = size(sequence,2);
end
%绘图前后排序的各组数据个数
figure
subplot(1,2,1)
bar(sequenceLengths)
ylim([0 30])
xlabel('Sequence')
ylabel('Length')
title('Sorted Data')
%按序列长度对测试数据进行排序
[sequenceLengths,idx] = sort(sequenceLengths);
XTrain = XTrain(idx);
YTrain = YTrain(idx);
subplot(1,2,2)
bar(sequenceLengths)
ylim([0 30])
xlabel('Sequence')
ylabel('Length')
title('Sorted Data')
%%