Over the last few years I’ve listened to quite a few podcasts on Data Science, Statistics and Machine Learning.
I’ve listed some of them here in the hope that people find them useful. In time I hope to do small blog posts with quick reviews on selected episodes that I’ve listened to and outline some specific lessons or thoughts I got out of it.
All of these podcasts are available in iTunes – the links below are to the supporting websites where often there is additional material, episode transcripts, references to academic papers and links to source code for various tools and projects.
This too is a work in progress – there are many people with a lot to say out there…
This one is my absolute favourite. These are in depth discussions about machine learning, statistics and Bayesian optimisation from Assoc. Prof Ryan Adams and Katherine Gorman. In the first few minutes of every episode Ryan talks about new developments in the industry and/or answers a question sent in by listeners. I simply sit there in disbelief at how well he describes complex concepts in such a casual manner.
This one, however, is not for the um…attention span challenged as they tend to average about 40 minutes per episode.
Presented by applied statistician Keith Bower as a series of short lectures (about 5 minutes a piece on average) on very specific topics in statistics. Available in either video or audio, these opened a number of areas of new discovery for me. Keith’s passion is contagious and his presentations are well work watching (the video episode on the Poisson distribution is gold). Very highly recommended.
I’m still working through this one as you really have to concentrate. Dr Golden is an absolute pleasure to listen to and I just love imagining this guy recording in his basement and have a great old time. Presented as a series of “How-to” episodes of 30-40 mins.
I like this one particularly because of the interaction between Ben who comes from a web design background and Katie who is a working astro-physicist. These average about 20 mins an episode.
A great one if you need to get into basic R. I struggled a bit as it was very basic however it does have some hidden gems in it. It also starts with some interesting history about how R came about (including why its called R in the first place). I’m still working my way through this one.
Husband and wife pair Kyle and Linda, this one alternates between 30-40 minute deep dive episodes with guests and short 10 minute mini-episodes where Kyle discusses a particular statistical/machine learning concept and Linda (who is non-technical) acts as his foil.
I have mixed feelings about how well this works so I tend to skip the mini episodes unless it’s regarding something I particularly want to reload into my working memory.