f



Cue/Readout network

Learns partial cues from the input and then learns a read-out layer to the wanted result.

https://drive.google.com/open?id=0BwsgMLjV0BnhUXNKcmVKdzFWTGM

Of course you can push it further and learn cues from cues.

I've used random projection neural nets but I think the idea fits in well the Hawkins/Ahmad viewpoint of how an individual neuron operates (but perhaps not their view about how groups of neurons operate.)

https://arxiv.org/abs/1511.00083
0
Sean
12/7/2016 5:23:15 AM
comp.ai.neural-nets 5773 articles. 2 followers. tomhoo (63) is leader. Post Follow

0 Replies
442 Views

Similar Articles

[PageSpeed] 49

Reply: