Training Neural Network: What does net.divideMode = 'value' mean?

I am wondering what the function net.divideMode = 'value'; means? Many people use it when prediction time series with NARX, but I cannot find anything in the web or documentation.

What is the difference between "value" and "time" ?

Thank you

[Greg Heath]

Have you compared the two on an example?

[Greg Heath]

My answer is I don't know. Any explanation that I have tried to understand didn't help.

I ran a quick narxnet comparison on the simpleseries_dataset. The results were the same.

So I guess a good question would be

When are the results different?

Good Luck.

P.S. If you find out, please post.

[Tim]

Yes that was also my result, in the end I did not find any difference between "time" and "value". But I am still wondering why people use it and where it comes from. Thank you anyways!

Answer 1

[Ghada Saleh]

Hi Tim,

The 'net.divideMode' is a property that defines the algorithms to use when a network is to adapt, is to be initialized, is to have its performance measured, or is to be trained.

This property defines the target data dimensions which to divide up when the data division function is called. Its default value is 'sample' for static networks and 'time' for dynamic networks. It may also be set to 'sampletime' to divide targets by both sample and timestep, 'all' to divide up targets by every scalar value, or 'none' to not divide up data at all (in which case all data is used for training, none for validation or testing).

I hope this helps,

Ghada

[AYUSHI AGRAWAL]

Hey, what if i dont want to use the divide method already inbuilt in the NN APP? I want to divide samples outside the APP. I tried doing that but somehow its not working at all. Does the division also follows normalizing the data?

For context, I am trying to make a neural network for a data with 2 inputs and 2 outputs. (I have used the Regression APP)