Hello,
when we fit a neural network model to a dataset and select the save formula option, the results are saved. For example, if I want to see the formula for H1_1, it is written as:
0.5 * ((-12.824088735563) + Match( :sex, 0, 6.99084931122274, 1, -(6.99084931122274 + 0) ) + -0.166405890007763 * :weight
+0.436280476305382 * :age + 2.18056058975802 * :tumor size + Match( :family, 0, -3.24492142132066, 1, -((-3.24492142132066) + 0) )
+Match( :hyper, 0, -0.695306127956332, 1, -((-0.695306127956332) + 0) ) + Match( :dissection,
0, 0.12239140444254,
1, -(0.12239140444254 + 0)
))
My question is, how are the coefficient numbers written in this formula obtained?
You might find more information about the estimation methods in this technical report.
I'm fairly sure you should find some sort of an answer from JMP Documentation/Help for Neural Networks: Neural Networks (jmp.com)
