Home 2023 › Forums › General Psychology › Suggestions on good Neural Network Resources?
Hi,
As I know we have a lot of educated members here, I am wondering if any of you have recommendations for good resources on neural networks?
I've reached a technical/language barrier, in describing metabolic 'computation', and I realize I need to read up on a lot more things before I'm ready to formulate these ideas further. Bonus points if the material intersects with cognitive science + philosophy of mind + programming.
(Actually, just send me anything related!)
(Thanks!)
I think cognitive architectures are related: https://en.wikipedia.org/wiki/Cognitive_architecture
Also: https://www.youtube.com/watch?v=bfO4EkoGh40
Check out SOAR, ACT-R, and SIGMA.
Lastly: https://cs.uwaterloo.ca/~jhoey/teaching/cogsci600/papers/laird.standard.aimag.2017.PDF
Don't know if you saw what I linked awhile back, but these are what I was talking about.
Here's some literature on cognitive architectures and personality specifically:
https://www.aaai.org/Papers/Symposia/Spring/2004/SS-04-02/SS04-02-017.pdf
https://www.dropbox.com/s/dt02n4qijz90i62/sun-COGSYS-2014.pdf?dl=0
This is what we've been using for computer science class:
http://neuralnetworksanddeeplearning.com/index.html
http://www.deeplearningbook.org/
http://colah.github.io/posts/2014-03-NN-Manifolds-Topology/
Playground to test networks on:
https://playground.tensorflow.org
Thank you so much for these guys, I've been digging into them and quite enjoying the learning experience. @supahprotist your introduction of the SOAR cognitive architecture is very interesting to me, as I see so many correlations to my recent computational metaphor thread. Here's a quick superimposed schema of it and CT:

What I called the "apparatus" in the computational metaphor thread appears to be essentially what this is ^ , so it's encouraging to know this has been worked on before. It gives me a good baseline from which to better articulate the architecture of CT in computational language, which is just what I was looking for.
(p.s. granted I don't think SOAR maps onto CT 100%, and I'm still reading about it. The labels may only be semi-related in the terminology employed, but 'semantic' seems to be about 'facts' which is not how Ji works at all. So there are key differences in that one at least)
Artificial or Biological Neural Networks?
.
.
.
