I have implemented an ANN with 2 hidden layers, activation PRelU. I have saved my model in an hdf5 file and am trying to interpret the results from model.save (kernel, bias, and alpha)... I can see the tables, heatmap, and charts of these. Now, how do I interpret them and understand which nodes are useful, do I need to change the bias etc.?
Tried reading many online resources, but haven't found anything very useful on what my next steps should be.[![Kernel[