References to learn brain science from CS background

Post Reply New Topic

I'd like to learn more about how the brain works, especially functions of different areas, how neurons build circuits to process sensory inputs, store memory, and so on. I'm intrigued by some of the research out there, and want to understand the details of things like the Intense World Theory.

Can anyone suggest any references short of going back to grad school? My background does include heavy computer science, machine learning, and math - but no wet science - so I am really hoping to find a source that makes reference of what I already know about artificial neural nets. It's tricky finding out what's the current state of knowledge - I could go read Marvin Minsky, but I think some of his ideas are now debunked? What's our current best understanding?

I know some of you out there do this all day for a living... any ideas? Thanks!

Read anything related to cognitive neuroscience--the more recently written, the better.

Also: read articles pertaining to current research in autism and other neurological disorders.

Also: read standard neurology textbooks--again, the more recently written, the better.

Here are a few journals that I recommend:

http://brain.oxfordjournals.org/

http://www.sciencedirect.com/science/journal/00068993

I hope they are useful to you.

Some books you can google are:
Theoretical neuroscience by Dayan & Abbott
Neural networks for pattern recognition by Bishop
Intro to computational neuroscience by I forgot the authors, but it is a classic and will appear on top of results

For basic intro to neuro, the best book is principles of neural science by kandel
If you are interested in vision including computational aspects, try vision science by palmer

Thanks everyone. That will keep me busy. Now if I only had a brain scanner at home...

You can create behavioral tasks on your computer and have people do them and computationally model mental processes. Maybe you will discover something new.

You might be able to explore some neuroimaging datasets that have been posted online: https://openfmri.org/

Cool. The computational neuroscience would be fun. For example, train an artificial neural net to be hyper-sensitive, hyper-plastic as described in Intense World Theory. I saw some journal articles about that. It might need to be prohibitively large.

I got interested reading Temple's book, where she describes some of her neuroimaging differences. Personally I think my amygdala is probably the size of a grapefruit.