这是你的answer . 唐't look at the accepted answer, but one with most votes (by Ark-kun). In general, multilayer perceptron with one or two (in case of nonlinear functions) is your way to go. Just remember NOT to add any regularisation to the network, as you don'吨有任何噪音 .
1 回答
这是你的answer . 唐't look at the accepted answer, but one with most votes (by Ark-kun). In general, multilayer perceptron with one or two (in case of nonlinear functions) is your way to go. Just remember NOT to add any regularisation to the network, as you don'吨有任何噪音 .