I'm working on a project where I need to perform both forecasting (regression) and classification using time series data. The dataset is labelled.

I've been exploring LSTM networks due to their effectiveness in handling sequential data.

Is it possible to use a single LSTM model for both forecasting and classification tasks?

  • $\begingroup$ What do you mean by a single model? Multi-Task learning is a good option, it would be a model with multiple outputs. $\endgroup$
    – Dr. Snoopy
    Oct 28, 2023 at 11:32
  • $\begingroup$ Do you mean something like both deciding if an image is of a dog, cat, or alligator (classification) AND how tall the animal is (regression)? $\endgroup$
    – Dave
    Oct 31, 2023 at 2:51
  • 1
    $\begingroup$ @Dave yes this is a similar case. $\endgroup$
    – mike7
    Nov 1, 2023 at 10:55


You must log in to answer this question.