Skip to main content

W07. State 03

Today we will learn how to apply Bayes Theorem in a chain to create Bayes Networks.

Pre-readings and Videos​

It may seem like detection should "just work" if you get the right threshold, but usually there's so much uncertainty in the measurement and error in the robot movement that we need to use a process to iterate towards a probable state. Iterating measurements allows us to converge to a detection.

Bayes Networks​

This description of Bayes networks gives a technical outline of Bayes Networks. It is worth reading through in some detail if you would like to use Bayes Networks to model inferences.

Bayes Networks for Diagnosis​

This video describes the use of a Bayes Network to perform a diagnosis of lung problems based on a series of known and unknown variables.

Summary of the Day​

Learning Goals​

  1. Be able to read a Bayes Network in terms of causality.
  2. Be able to understand the propogation of an observation throughout a network.
  3. Be able to create a simple Bayes Network to model causality and observations.