Model from data and optimization

Hi everyone, I have a set of data from FEM simulations, I need to generate a mathematical model from them and after that find optimal value for at less 9 different variables. I am looking at optimization problems but looks like this aproach first requiere knowing the equation or the relationship between variables, and for neurnal network aproach, after trying the model I will need to put the values for the variables to predict the outcome, but is not the idea. I will really appreciate any help, thanks.

Answer 1

[Shashank Gupta]

Hi Juan,

Seems like a prediction model, you can use any neural network framework that fits your data appropriately and used the model as mathematical framework and subsequently predict the outcome. That is certainly not the only way, how about you give a try to polynomial equation and fit your data, check out this. The idea to approach this problem is to get some mathematical model that fits your data. try using this resources link.

I hope this helps.

Cheers

[Juan Pablo Guamn Bernal]

great, thank you so much