I am currently working on a project on Long Short Term Memory for Sequence to Sequence and Sequence to label Classification on matlab.
I am using excel data set.
data = xlsread('energydata.xls');
XTrain=data(1:19735,1:1);%input data set
%YTrain=data(1:19735,28); %target data set
Please how can I convert the excel data set to cell aray and categorical arrays?
I might have a look at the following thread:
https://www.mathworks.com/matlabcentral/answers/307835-how-can-i-import-excel-data-in-a-cell-array
hello Thomas,
if you have import data from excel, and now you had a matrix training data and testing data, next you can convert into cell array using function mat2cell()
example for the training data
XTrainArray=mat2cell(XTrain,ones(length(XTrain),1))
i hope it helps
Best Regards,
Hi Thomas,
I've done it for sequence-to-regression LSTM. But I think it works for you too.
You should have several sequences in cell format and also have their corresponding labels.
Just read data from excel file and use a loop like this:
period=70;
for i=1:size(data,1)-period
seq{i}=SequenceInExcel(i:i+period-1,:);
Y(i)=Label(i);
end
For me, it was necessary to use sequences with a length of 70. You can change it. Actually you don't have to have equal-length sequences. But it seems you should have several sequences in cell matrices. After that, you can train the LSTM like this:
net = trainNetwork(seq,Y',layers,options);
- Badges
- Science topic