I have data in a table, it has 5 variables. The data is obtained by solving a nonlinear equation at each point.
I would like to use it to make some sort of analytical approximation function for $t$ based on variables $a,b,c,d$: $$t = f(a,b,c,d) $$
I am wondering if there is a way to do this with neural networks?
another thing I have looked into is gradient descent (not sure if thats limited to 4 variables)
How could I implement this in MATLAB?