Is a critical tool for avoiding errors while working with skewed data. Assuming an event A, such as a percentage of patients with cancer, or the number of actual drug users in a population. And an event B which is a positive test result for cancer, drug use, whatever... If the test is positive when the actual event has occurred some percentage of the time p(B|A), and negative when it should be by some percentage p(¬B|¬A) aka the opposite of it being positive when it should not be p(B|¬A), Bayes Rule tells us what the actual chance that the event A has happened give a positive test result B.

` p(A|B) = ( p(B|A) * p(A) ) / p(B)`

Where A and B are events. p(A) and p(B) are the probabilities of those events.
p(B|A) is the probability of seeing B if A is true. p(B) can be calculated
as `p(B|A) * p(A) + p(B|¬A) *
p(¬A)` where
¬ denotes NOT. e.g.
p(¬A) = 1 - p(A)

Also:

- Machine Learning Method, Bayesian Classification
- Troubleshooting Machine Learning Methods
- Localization (computing position over time from sensor readings)

file: /Techref/method/ai/bayesrule.htm, 2KB, , updated: 2017/4/19 13:58, local time: 2024/8/4 23:21, |

©2024 These pages are served without commercial sponsorship. (No popup ads, etc...).Bandwidth abuse increases hosting cost forcing sponsorship or shutdown. This server aggressively defends against automated copying for any reason including offline viewing, duplication, etc... Please respect this requirement and DO NOT RIP THIS SITE. Questions?<A HREF="http://www.piclist.com/Techref/method/ai/bayesrule.htm"> Bayes Rule</A> |

Did you find what you needed? |

PICList 2024 contributors:
o List host: MIT, Site host massmind.org, Top posters @none found - Page Editors: James Newton, David Cary, and YOU!
* Roman Black of Black Robotics donates from sales of Linistep stepper controller kits. * Ashley Roll of Digital Nemesis donates from sales of RCL-1 RS232 to TTL converters. * Monthly Subscribers: Gregg Rew. on-going support is MOST appreciated!
* Contributors: Richard Seriani, Sr. |

## Welcome to www.piclist.com! |

.