How to create dataset from my tribological experimental data?

I have a doubt in using ANN. I have some experimental results with me. There is a total of 27 experiments (L27 array). Three Inputs Load (3 level), frequency (3 level) and weight % (4 level). There are two outputs viz., Coefficient of friction and Specific wear rate. I need to predict the outputs for the input levels beyond those values which I have conducted the experiment. Can you please guide me to train a successful network for correct prediction using any advance tool in MATLAB 2023a? I have no idea selecting no. of hidden layers,no of neurons, which transfer function, epochs, mu,mu_fail, goal etc etc. Please support me in this problem.

[Shreeya]

Hi

Follow the example here for designing a neural network with multiple outputs. Network parameters such as number of hidden layers, epochs etc depend on the data being trained.

Hope this helps.

Answer 1

[Rohit Kulkarni]

Hi,

I understand that you want to train a neural network with multiple inputs and outputs, you can refer to the following documentation:

Multiple-Input and Multiple-Output Networks - MATLAB & Simulink - MathWorks India.

Also refer to this MATLAB answer.

Thanks