good day

I have a problem where i have 9 data points that are collected every minute for 40 minutes, and by the 40th minute the solution would be either end up being black or white.

I would like to set up a neural network; which would take the live input of every minute; and I was hoping within the 25-30 minute mark to predict the outcome of what the results would be at 40 minute; which is a classification.

I have over 3000 historical runs of this experiment; each containing 40 rows of 9 columns data per experiment.

What network would I need to set up; so that it can learn from each run at every minute mark per experiment with the results; and then set it up for live input, when the experiment is running again.

i feel like I might need more than one system to accomplish this; any help in pointing me towards the right path would be greatly appreciated

I am using python (keras) to try to solve this problem.


So it seems any time series modeling would do the trick. If you new to neural nets and seems you want to play with keras, maybe start by throwing it into a simple LSTM.

  • $\begingroup$ thank you for your answer; I will investigate LSTM networks. Much appreciated $\endgroup$
    – Rad D
    May 31 '19 at 19:50

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.