This makes my brain hurt, but in a nice way. It’s seems like the other end of machine learning. Interesting times. My EE trained brain found this overview particularly straight forward.
This makes my brain hurt, but in a nice way. It’s seems like the other end of machine learning. Interesting times. My EE trained brain found this overview particularly straight forward.
Some of the videos might help as well:
http://igorcarron.googlepages.com/csvideos
Cheers,
Igor.
Igor – Thanks!
Those are fun; the one from Columbia where they recover a video of milk dispersing in an aquarium is cool. But it’s interesting that I had read all that other stuff and not entirely appreciated the importance of having your sampling be “incoherent” v.s. your the family of functions your fitting to the data. A requirement that is, it appears, implemented by random sampling. That’s wonderful, on the recreation brain hurting scale.
This incoherence property is indeed interesting and most people including me are still in awe when seeing that it works. It goes against most of our upbringing as engineers.
Igor.