Model Identification with Matrices

Hello, I have the following measured data (no time sequence !):

A = [30x1] %input 1
B = [1x5} %input 2
C = [30x5] %output

So the model can be described by:

A * B = C

For identification I look for a function like:

C = f(A, B)

There are no parameters required for A or B, because this is measured data.

What is the best way to identify such function ? Grey-Box-Modelling / Least-Square ?

Can you post a code with small explanation please ?

Thank you very much

[Marv]

No Idea ?