Hi, is it possible to give more instructions or explanation on the phrase “the predications can be traced”? What should we specify if we use neural network ? thanks ~
Hi
This is a topic linked to explainability: you just need to be able to explain your results, i.e. to show what factors drive your predictions.
In your NN case, you could check which neurons fire up at each layer for instance, or which activations have the high values.