Prerequisites: Graduate standing, three semesters of calculus, knowledge of programming. Additional information is available here.
Instructor: Dr. Anthony Maida.
New: 4/8/09 Useful code snippets
Here is MatLab code to show how to plot uncertainty ellipses from two-dimensional covariance matrices.
Here is example matlab code to implement the velocity model on page 210 of the textbook. Use it as a reference to translate into Java.
A library for matrix manipulation in Java is available here.
The text warns (p. 135) that computing the difference between two angles is tricky because the result must always be normalized to the range [-pi , pi]. A function to do that is given here.
Here is Java code that shows how to write data to a file for subsequent plotting by MatLab. Here is a MatLab script that shows how to plot the data.
New: 2/19/09
Here are sample questions to study for the first midterm. The first midterm is Thursday Feb 26.
New: 4/21/08
Here is corrected MatLab code to show how to plot uncertainty ellipses from two-dimensional covariance matrices.
New: 4/8/08
Here is Java code that shows how to write data to a file. Here is a MatLab script that shows how to plot the data.
New: 4/3/08
Starter code for the class project is available here in zip format. Documentation for the code is here. An ANT build file for the project is here
| Teaching assistant: | Padraic Edgington | |
| Email: | padraic@louisiana.edu | |
| Phone: | (337) 482-1696 | |
| Office: | ACTR, Room 331 | |
| Office hours: | Mon. | 2pm-4pm |
| Tues. | 2pm-4pm | |
| Thurs. | 2pm-4pm |
| Requirements and grading: | |
| 1st Midterm: | 25% |
| 2nd Midterm: | 25% |
| 3rd Midterm: | 25% |
| Programming project: | 25% |
| Textbook: |
| Probabilistic Robotics |
| Sebastian Thrun, Wolfram Burgard, Dieter Fox |
| MIT Press, 2005 |
| ISBN: 0-262-20162-3 |
The first draft of the class project specifications is available here.
A library for matrix manipulation in Java is available here.
The text warns (p. 135) that computing the difference between two angles is tricky because the result must always be normalized to the range [-pi , pi]. A function to do that is given here.
Supporting material for the Probabilistic Robotics Text.
Sutton and Barto's online book on reinforcement learning.
This paper, written by Thomas Dean, titled "Computational Models of the Cerebral Cortex," advocates the use of hierarchical Bayesian networks to model the brain.
This paper, written by Tai Sing Lee and David Mumford, titled "Hierarchical Bayesian Inference in the Visual Cortex," advocates the use of hierarchical Bayesian networks to model the visual areas of the brain. They specifically advocate the use of particle filters.