I am trying to dig deeper into the Neural Network algorythims. I have found a couple of posts talking about setting parameters but I must be doing it wrong because I never get any difference in results.
Here is what I am trying:
c = Predict[association,
Method -> {"NeuralNetwork", "HidddenLayers" -> {10, "Linear"},
"NeuronTypes" -> {"Linear", "Linear"},
"NumberOfNodes" -> {10, 5, 1}}, PerformanceGoal -> "Quality"]
When I change these parameters and then check them with:
Normal[Options[c]][[1]][[6]]
I get the exact same results no matter how I tweak the starting parameters. By results I mean, actual prediction results and also parameter settings. Well, actually, one exception is:
"MaxIterations" -> 900
Sometimes that single value changes. However, the rest do not.
Can someone enlighten me on how to properly tweak Neural Networks? I have read as much as possible and watched all the training videos.
Please and Thank You!