Revision as of 10:41, 4 September 2020 edit 2001:16b8:403a:3900:10ac:5987:da21:4c5 (talk) →Arbitrary Width Case ← Previous edit		Revision as of 11:12, 4 September 2020 edit undo 2001:16b8:403a:3900:10ac:5987:da21:4c5 (talk) No edit summary Tag: Manual revert Next edit →
Line 67: ::::::::: OK, sounds good. [[User:Koertefa\|''<span style="color:#2F4F4F">'''K'''<span style="color:Teal">œrte</span>'''F'''</span><span style="color:Teal">a</span>'']] [[User talk:Koertefa#top\|<span style="color:#2F4F4F">'''{'''<i style="color:Teal">ταλκ</i>'''}'''</span>]] 19:22, 6 July 2020 (UTC) ~~== Arbitrary Width Case ==~~ I am wondering if the theorem presented here is correct. In http://www2.math.technion.ac.il/~pinkus/papers/acta.pdf, the neural network has only one node in the output layer (so they map to R). Here the number of output nodes is arbitrary. Is this correct?

Talk:Universal approximation theorem: Difference between revisions