Thomas Bayes

mathematician · 9 mentions across 3 readings

In this course

Bayes is invoked here primarily through the problem of "Bayes' circle"—the epistemic trap where our posterior beliefs filter how we measure the priors and likelihoods that supposedly justify those beliefs. The readings use Bayes' theorem as a framework for thinking about perception and inference, but are deeply concerned with whether Bayesian approaches can escape circularity when dealing with interface-world relationships and the verification of priors. The excerpts suggest that while Bayesian conditional probability remains theoretically productive for perception and detection, the course grapples with the methodological challenge of grounding Bayesian inference without begging the question.

Mentioned in 3 readings

Donald Hoffman

The Interface Theory of Perception

Bayes’ Circle: We can only see the world through our posteriors. When we measure priors and likelihoods in the world, our measurements are necessarily filtered through our posteriors. Using our measurements of priors and likelihoods to just…It is, of course, possible to pursue a Bayesian approach to perception without getting mired in Bayes’ circle. Indeed, Bayesian approaches are among the most promising in the field. Conditional probabilities turn up everywhere in perception…What might those options be? That depends on which constraints one postulates between interface and world. Suppose, for instance, that one wants a minimal constraint that allows probabilities of interface events to be informative about prob…

+1 more

I.J. Good

Speculations Concerning the First Ultraintelligent Machine

The rational procedure in perception would be to estimate the final (a posteriori) probabilities by means of Bayes’ theorem, and then perhaps to select one or more hypotheses for future consideration or action, by allowing also for the util…If the index terms $W1, W2, \dots, Wm$ provide statistically independent information concerning $Dj$ (i.e., if $W1, \dots, Wm$ are statistically independent, and are also statistically independent given $Dj$), then $$\log P(Dj \mid W1 \cdo…which have been used, for example, in the simplest class of perceptra [82], and in suggestions or experiments related to mechanical chess and chequers (draughts) (for example [33, 83, 18, 83a]). One can write (5.4) in the form $$aj + \sum…

+1 more

Finn Brunton and Helen Nissenbaum

from Obfuscation: A User’s Guide for Privacy and Protest

“Rather than trying to hide information from Facebook,” Ludlow wrote, “it may be possible simply to overwhelm it with too much information.” Ludlow’s experiment (which he called “Bayesian flooding,” after a form of statistical analysis) en…

Appears alongside

People mentioned in the same passages — sorted by co-occurrence weight.

Donald MacKay 1 Franz Joseph Gall 1 Friedrich Hayek 1 Middleton 1 William McDougall 1 Marvin Minsky 1 Stephen Palmer 1 Peter Ludlow 1