Cortical Learning VS Deep Learning


#1

hi,
I read some papers and I find this comparaison :


I have two questions
in this picture whay CL is not ML ?
and the CL is a ANN or not ?
well I know the answers but after seeing this I’m confused


#2

IMO, The paper is talking about CL exclusively regarding real brains built with living cells. Instead of ML algorithms that mimics the brain (HTM, SPAWN, etc ). While most paper defines ANNs as Neural Networks that is trained using backpropergation.


#3

While most paper defines ANNs as Neural Networks that is trained using backpropergation.

but the HTM and the DL use the same notion (neurons and layers), but in a different way. the DL based on the mathematical function and the activation functions but the HTM is based on the biological method.


#4

They are very different things even they share the same abstraction! C++ and Java are very different despite they share a similar syntax.
Besides, no one talks about proximal/distal connections in Neural Networks while no one talks about tensors and operators in HTM! That’s a big enough difference, isn’t it?


#5

yes of course, they are very different. It istruth that the CL and the DL use the neuron. but they work differently, thanks marty ^^