1
$\begingroup$

I have 2 binary outputs (1 and 0) with time series data. The dataset order is shown in the image.image.Can anyone suggest me how to handle this problem with LSTM? Particularly in MATLAB or Python. Thanks in advance.

$\endgroup$
1
$\begingroup$

This can be done with RNN/LSTM/GRU (type of Neural Networks that are well-suited for time-series).

For example : https://machinelearningmastery.com/multivariate-time-series-forecasting-lstms-keras/

This example is quite similar to the problem mentioned in question (predict air quality based on ~10 parameters. Parameters are available as a time-series).

$\endgroup$
2
  • $\begingroup$ @shamit..Thank you for a useful answer.....What if i remove the time and date coloumn and implement in LSTM?.Is it a good idea to remove time series and apply remaining inputs to LSTM? $\endgroup$
    – Case Msee
    Feb 7 '19 at 0:47
  • $\begingroup$ Yes, do try that version (data minus time / date columns). Also try creation of derived columns (Like time of day, month) to capture seasonality in data (Like temperature in month of May @ 11 AM vs 11 AM in Dec. $\endgroup$ Feb 7 '19 at 2:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.