Hi dear researchers,
I want to calculate sensitivity of Neural Network inputs, for example if it has 10 inputs, I want to know for example how much input number 1 effects in holl Network? What is sensitivity of other inputs as well? Could you please give me some notes. Related formula will be quite helpful.
Thanks
I think the answer to your question is the weights associated with your required input . The more the weights the more it influence the network.
Dear Fatameh;
Your question is quite close and needed to be described more. In a monte carlo simulation, you can have different numbers of outputs. On the other hand, those ourputs can be each differet one input for an ANN approach. I can strongly offer that, first describe your potential expectations from ANN and revise your inputs according to them. Otherwise, you can not decide for a healthy relationship between MC outputs and ANN inputs"
Moreover, There is no rule or formula for this relationship. Only thing that you have to make your balance for outputs from MC and arrange them as input file for ANN.
I have some publications that you can read in my personal ResearchGate profile regarding to MC and ANN.
- Badges
- Science topic