Comments on: Neural Network “Learning Rules” http://www.neurevolution.net/2007/03/15/neural-network-learning-rules/ Chronicling the cognitive revolution in neuroscience Sun, 05 Aug 2012 00:06:50 +0000 hourly 1 http://wordpress.org/?v=3.0.4 By: Neurevolution » Blog Archive » A Brief Introduction to Reinforcement Learning http://www.neurevolution.net/2007/03/15/neural-network-learning-rules/#comment-13786 Neurevolution » Blog Archive » A Brief Introduction to Reinforcement Learning Mon, 02 Jun 2008 15:36:33 +0000 http://www.neurevolution.net/2007/03/15/neural-network-learning-rules/#comment-13786 [...] of learning hypothesized to occur in various parts of the brain, i.e., Hebbian-like LTP (here and here) and error-correction learning (here and [...] [...] of learning hypothesized to occur in various parts of the brain, i.e., Hebbian-like LTP (here and here) and error-correction learning (here and [...]

]]> By: Neurevolution » Blog Archive » Levels of Analysis and Emergence: The Neural Basis of Memory http://www.neurevolution.net/2007/03/15/neural-network-learning-rules/#comment-13643 Neurevolution » Blog Archive » Levels of Analysis and Emergence: The Neural Basis of Memory Fri, 30 May 2008 15:45:04 +0000 http://www.neurevolution.net/2007/03/15/neural-network-learning-rules/#comment-13643 [...] Raymond and Redman replicate the earlier finding that longer bouts of electrical stimulation can cause LTP to be more powerful (resulting in larger postsynaptic responses), and last longer.  They demonstrated three different levels of LTP in their experiment by using three different length trains of electrical stimulation.  This stimulation-dependent property of LTP has been taken as the basis for synaptic modification rules used in neural network models (http://www.neurevolution.net/2007/03/15/neural-network-learning-rules/). [...] [...] Raymond and Redman replicate the earlier finding that longer bouts of electrical stimulation can cause LTP to be more powerful (resulting in larger postsynaptic responses), and last longer.  They demonstrated three different levels of LTP in their experiment by using three different length trains of electrical stimulation.  This stimulation-dependent property of LTP has been taken as the basis for synaptic modification rules used in neural network models (http://www.neurevolution.net/2007/03/15/neural-network-learning-rules/). [...]

]]> By: Seth Herd http://www.neurevolution.net/2007/03/15/neural-network-learning-rules/#comment-4794 Seth Herd Wed, 07 Nov 2007 06:10:03 +0000 http://www.neurevolution.net/2007/03/15/neural-network-learning-rules/#comment-4794 Excellent post, and excellent paper! I actually read your paper when it came out. As best as I can tell, very few people are thinking about temporal sequencing in neural nets. Excellent post, and excellent paper! I actually read your paper when it came out. As best as I can tell, very few people are thinking about temporal sequencing in neural nets.

]]> By: Chris Chatham http://www.neurevolution.net/2007/03/15/neural-network-learning-rules/#comment-313 Chris Chatham Wed, 28 Mar 2007 01:01:55 +0000 http://www.neurevolution.net/2007/03/15/neural-network-learning-rules/#comment-313 Wow, this post is awesome. I have a similar (but much less clear) set of notes that I review every time I open up the simulation software I use! This is much better. This is my first time at this blog - very cool stuff! I'll be adding you to my blogroll. Wow, this post is awesome. I have a similar (but much less clear) set of notes that I review every time I open up the simulation software I use! This is much better.

This is my first time at this blog – very cool stuff! I’ll be adding you to my blogroll.

]]> By: P.L. http://www.neurevolution.net/2007/03/15/neural-network-learning-rules/#comment-134 P.L. Fri, 16 Mar 2007 01:30:42 +0000 http://www.neurevolution.net/2007/03/15/neural-network-learning-rules/#comment-134 No you've got it right -- to make sure to not confuse the newbies, and to make it as clear as possible, I'll make that change now! Thanks! No you’ve got it right — to make sure to not confuse the newbies, and to make it as clear as possible, I’ll make that change now! Thanks!

]]> By: Eric http://www.neurevolution.net/2007/03/15/neural-network-learning-rules/#comment-133 Eric Fri, 16 Mar 2007 01:14:02 +0000 http://www.neurevolution.net/2007/03/15/neural-network-learning-rules/#comment-133 Thank you. Very interesting. Just one point on your explanation of the formula - in your example if W(now) = 0.5 then you should use consistently throughout your formula, otherwise a newbie like me will get confused. W=0.5 + LearningRate x [PRE(past)-0.5]... Unless I am wrong? Thank you. Very interesting. Just one point on your explanation of the formula – in your example if W(now) = 0.5 then you should use consistently throughout your formula, otherwise a newbie like me will get confused. W=0.5 + LearningRate x [PRE(past)-0.5]… Unless I am wrong?

]]> By: Shane http://www.neurevolution.net/2007/03/15/neural-network-learning-rules/#comment-132 Shane Thu, 15 Mar 2007 23:01:26 +0000 http://www.neurevolution.net/2007/03/15/neural-network-learning-rules/#comment-132 Awesome article. Thanks. Awesome article. Thanks.

]]>