Comment on General Theory of Neural NetworksparentComments−rdlecler1OP2yThe topology needs to be information bearing. Weights of 0.0001 are likely spurious and if other weights are so relatively big they can effectively make the other fan in weights spurious as well.
Comments
The topology needs to be information bearing. Weights of 0.0001 are likely spurious and if other weights are so relatively big they can effectively make the other fan in weights spurious as well.