What is the difference between PredictionESN and RegressionESN when we predict a time series through it's factors?

[imhithanks]

I would like to predict a time series Y through using It's factor X1 and X2.
X1and X2 are time series. Y relates to X1 ,X2 and Y's history values .
I regard X1 and X2 as inputs. Y is output.

1. What is the difference between PredictionESN and RegressionESN？
2. Which one should I choose to predict Y through using X1, X2 and Y's history values
   Thank you so much!

[zimmerrol]

Hi,
the RegressionESN class takes a time series as input and produces a stationary, not time-dependent output. Example: take a sine wave and let the ESN infer the frequency of the sine wave.
The PredictionESN on the other hand takes a time series as input and infers the next step of said time series. Example: you have values of a wave function for x € [0, 5] and want to infer how the function will look like in the range x € [5, 10].
Does that answer your question?

[zimmerrol]

Closing as there is no activity anymore.