Gaussian Process Basics

author: David MacKay, University of Cambridge
published: Feb. 25, 2007,   recorded: June 2006,   views: 210532


Related Open Educational Resources

Related content

Report a problem or upload files

If you have found a problem with this lecture or would like to send us extra material, articles, exercises, etc., please use our ticket system to describe your request and upload the data.
Enter your e-mail into the 'Cc' field, and we will keep you updated with your request's status.
Lecture popularity: You need to login to cast your vote.


How on earth can a plain old Gaussian distribution be useful for sophisticated regression and machine learning tasks?

See Also:

Download slides icon Download slides: gpip06_mackay_gpb_01.pdf (3.6 MB)

Help icon Streaming Video Help

Link this page

Would you like to put a link to this lecture on your homepage?
Go ahead! Copy the HTML snippet !

Reviews and comments:

Comment1 Rohan Anil, December 31, 2008 at 1:15 a.m.:

Amazing, must watch
crystal clear!

Comment2 swy, June 11, 2009 at noon:

Video is very good, but can we not use rtmp? I would like to save the video so that I can watch it even when I am offline.

Comment3 zslevi, July 28, 2009 at 11:27 a.m.:

So, is the relation between a covariance and inverse covariance like between posterior and prior distributions?

Comment4 Lee, November 5, 2009 at 12:26 a.m.:

video is good, audio is poor though, hard to understand the speaker. My kingdom for a mic!

Comment5 Flávio Coelho, December 2, 2009 at 3:51 p.m.:

What software is he using to generate the graphics in real-time?

Comment6 gordon anderson, February 1, 2010 at 3:38 a.m.:

Very engaging and intuitive intro to gaussian processes.

Highly recommended.

Comment7 Tim, July 9, 2011 at 4:43 p.m.:

Excellent lecture, spoilt by shoddy camera work!

Comment8 human mathematics, January 9, 2012 at 9:57 p.m.:

A little hard to follow because he goes along without explaining why or how.

If the lecturer is reading this--I would use a ggplot geom_smooth / geom_path sort of deal rather than the video. The video is not Tufte-compliant because it shows the method more than the data.

For those having trouble viewing the videos, like me, stream the mms:// url with mplayer, that worked better for me than the site. I used mplayer -vfm ffmpeg -lavdopts lowres=1:fast:skiploopfilter=all mms://

Comment9 Louis, March 23, 2015 at 5:02 p.m.:

The RTMP link is rtmp://

If that stops working see

Comment10 Jarv, November 1, 2017 at 10 a.m.:

youtube-dl is capable (2017-10) of extracting this video. The RTMP links appear to be down.

The 'medium bandwidth' .flv stream appears to end up with the best quality (VP6 Flash codec), 320x240:

youtube-dl -v -xk -f http-415 ""

List available formats with:
youtube-dl -v -F ""

Comment11, January 11, 2021 at 5:31 a.m.:

A Gaussian process is a probability distribution over possible functions that fit a set of points.

Write your own review or comment:

make sure you have javascript enabled or clear this field: