nueral_network forward propagation no activation

[yummyyummya]

in neural_network_raw, the forward propagation function neural_net(x), layer_1 = Wx+b, but there is no activation sigmoid(layer_1), this will bring some problems.

[vishalmhjn]

yes, it is surprising how can this NN work without activation functions

[DCNick3]

Well, its accuracy is same as logistic regression. Adding activation makes it much more accurate.