[Computer-go] Fwd: Representing Komi for neural network
hughperkins at gmail.com
Fri Mar 20 19:08:53 PDT 2015
> But then, the komi wont really
participate in the hierarchical representation we are hoping that the
network will build, that I suppose we are hoping is the key to
obtaining human-comparable results?
Well... it seems that Hinton, in his dropout paper
http://arxiv.org/pdf/1207.0580.pdf , get kindof ok results with
'permutation-invariant' networks, basically consisting of 2-3 fc
layers, so maybe a bunch of conv layers feeding into 2-3 fc layers,
and the non-image inputs going into the fc-layers too is reasonable.
Perhaps what we want is a compromise between convnets and fcs though?
ie, either take an fc and make it a bit more sparse, and / or take an
fc and randomly link sets of weights together???
More information about the Computer-go