ISCA: PhilosophyThe ISCABBS forum: Philosophy(EarthQuaker30557) - Faunus: that may even be too harsh a judgement. They have developed some- June 11, 2008 Faunus: that may even be too harsh a judgement. They have developed some amazing toolsets for the analysis of neurological data. I don&39;t think Bak inc. ever did that much with self organized criticality.http://rss.iscabbs.com/getpost.cgi?forum=26&post=30557 (Faunus30556) - So the Bayesians are the neurological Per Bak :)- June 7, 2008 So the Bayesians are the neurological Per Bak :)http://rss.iscabbs.com/getpost.cgi?forum=26&post=30556 (EarthQuaker30555) - Faunus: that sounds about right. I haven't worked in sensory systems in a few- June 6, 2008 Faunus: that sounds about right. I haven&39;t worked in sensory systems in a few years now. Giraffe: Most of the application of Bayesian approaches has been at the level of very focal, control systems. For example, arm trajectory estimation, object motion tracking. What gets people riled at meetings (those I&39;ve been to, anyway) are the Reverend Bayes crowd who seem to assume that because a particular system is efficiently modelled as a Bayesian system, it is employing an...http://rss.iscabbs.com/getpost.cgi?forum=26&post=30555 (Plaid Rabbit30554) - Bayesian models of decision making have had mixed results. Mostly, we're not- June 6, 2008 Bayesian models of decision making have had mixed results. Mostly, we&39;re not very Bayesian behaviorally. But our visual system looks pretty Bayesian.http://rss.iscabbs.com/getpost.cgi?forum=26&post=30554 (Giraffe30553) - I don't know if the brain operates in a Bayesian manner or not. But I suspect- June 5, 2008 I don&39;t know if the brain operates in a Bayesian manner or not. But I suspect one could use Bayesian methods to make a pretty decent model of the way decisions get made. Does that come down to the same thinghttp://rss.iscabbs.com/getpost.cgi?forum=26&post=30553 (Faunus30552) - EQ: you heard about this discovery in vision, that says that our visual system- June 4, 2008 EQ: you heard about this discovery in vision, that says that our visual system directly compensates for the 110 second delay between retina and brain, so that what we are seeing is the scene which is presented to the retina, but the scene which our visual system predicts will exist 110 of a second in the future of what hit the retina I phrased that really badly but it&39;s hard to phrase well. Basically: what we are seeing at any given time is a prediction of what is out there right now,...http://rss.iscabbs.com/getpost.cgi?forum=26&post=30552 (EarthQuaker30551) - Vanity: yes, for example, how much information is transmitted per single- June 4, 2008 Vanity: yes, for example, how much information is transmitted per single action potential. It&39;s possible to optimize a code based on the theoretical upper limit of this information.http://rss.iscabbs.com/getpost.cgi?forum=26&post=30551 (Vanity30550) - "Optimal"- June 4, 2008 "Optimal"http://rss.iscabbs.com/getpost.cgi?forum=26&post=30550 (EarthQuaker30549) - The article isn't real good at explaining Bayes' Law, and I'm too tired to do- June 4, 2008 The article isn&39;t real good at explaining Bayes&39; Law, and I&39;m too tired to do a good job now. From my own work I find that Bayesian approaches are often the best approach to use for decoding neural signals...a lot of labs use that result to make the argument that Bayesian analytic techniques must be employed by cortical circuitry, as one would assume an optimal decoding scheme is operating inside the context of the brain itself. There is a hint of tautology here. However, there just.http://rss.iscabbs.com/getpost.cgi?forum=26&post=30549 (Faunus30548) - I have an idea. Let's arbitrarily assign a probability of 50% to the- June 3, 2008 I have an idea. Let&39;s arbitrarily assign a probability of 50% to the proposition that the brain is a big bayesian probability machine, and then adjust that value as new information comes in.http://rss.iscabbs.com/getpost.cgi?forum=26&post=30548 |