Recurrent Neural Networks for Demand Prediction - > Multi-Regression Approach?

Hello all,

I have just worked my way through the book “Codeless Deep Learning”.

In Chapter 6 Recurrent Neural Networks for Demand Prediction the author built a RNN to make prediction on energy consumption based on times series data. The author has mentioned that it is also possible to use past values from other external time series to make the prediction. So that we get a multi-regression approach. Unfortunately in the book the author does not explain this procedure in detail.

How would a RNN look like if I want to use other time series data, too? Are there any example workflows available which deal with multi-regression demand prediction with RNN?

Do I just concatenate single time series data in one table and execute the Network Learner node once? Or should I build a RNN for each time series and then somehow combine the results?

As you may notice, I am a complete beginner on this topic…So any help would be highly appreciated!!!

Thank you very much for your support in advance!

Hi @nicolai -

This blog post (by @Kathrin) just appeared today and I think addresses your question:

Maybe take a look and followup here with additional questions?