3 ms·
From the summary, they’re not altering the topology of the network at all, just the connection strength between neurons. While this isn’t the way a natural neu
by mr_toad 9y ago
From the summary, they’re not altering the topology of the network at all, just the connection strength between neurons.
While this isn’t the way a natural neural network would learn (I’m not sure a nematode can learn), it’s still interesting that you can take a copy of a natural neural network and force it to learn in this way.
- lurquer 9y agoI would expect there is an enormous set of random topologies that could be trained to balance a pole. Indeed, part of the elegant 'magic' of neural nets is that the topology is fairly irrelevant... the number of layers, the number of nodes, the manner in which they are connected... pretty much any configuration can get you into a mid 90% accuracy on MNIST (and emulating a basic PID algorithm is simpler than MNIST). Of course, I'm referring to basic tasks; clearly topologies matter a great deal with more sophisticated things.