-1
$\begingroup$

Lets say we have 2 data sets. First is the close price time series data set and we want to predict future values of it. The second is volumes of each price from the first data set and we do not want to predict on it but use this data to help predict future price values.

What kind of neural networks is suitable for this task?

In my sight this may be an LSTM with some changes.

Advise me please.

$\endgroup$
5
  • $\begingroup$ Hello There! May I ask why you want to use NN ? is your goal to do accurate predictions, or to just find a use for neural networks ? Depending on your feedback, we might recommend various algorithms for forecasting, that are not necessarily based on a neural network architecture. cheers! $\endgroup$ – Wajdi Apr 21 at 14:30
  • $\begingroup$ You can use RNN architectures like LSTM, and GRU. RNNs take input vectors in each time step, so you can add your extra data to input vector. Your input shape will be batch_size x sequence length x num_of_features. $\endgroup$ – tkarahan Apr 21 at 14:33
  • $\begingroup$ @tkarahan You should convert your comment to a formal answer. $\endgroup$ – 10xAI Apr 21 at 15:46
  • $\begingroup$ @10xAI I posted content of my comment as an answer. $\endgroup$ – tkarahan Apr 21 at 17:52
  • $\begingroup$ Hello @Wajdi! I believe that NN is more perspective but you have intrigued me. $\endgroup$ – Kroll Apr 26 at 11:39
1
$\begingroup$

You can use RNN architectures like LSTM, and GRU. RNNs take input vectors in each time step, so you can add your extra data to input vector. Your input shape will be batch_size x sequence length x num_of_features.

$\endgroup$
1
  • $\begingroup$ What if I have many-to-many with main sequence discrete valued [+1, 0 ,-1]? $\endgroup$ – Kroll Apr 26 at 11:33

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.