How to build a regression neural network from a datastore? How do i find the variable names from a datastore?

I have a combined file datastore with 5 predictor variables in 1 column and 1 objective variable in other column and row size of roughly 65000. I want to know the variable names so that i can use it to build a regression neural network. But i dont find a object function for knowing the variable names.

Also how do i use the datastore to build a regression neural network

Answer 1

[dpb]

'Pends on how you created the datastore and what it was built from...if you smooshed all five variables into one column, then you probably lost the variables' indentities in doing that and are left with only one.

You would want to build the datastore maintaining the variables, not merging them.

A 5-variable, 6500-length array is nothing as far as memory; there's no need for a datastore here.

tmp=rand(6500,5);
whos tmp

  Name         Size             Bytes  Class     Attributes

  tmp       6500x5             260000  double              

That's only a 260K array...