I read this paper about B-spline and a simple way to adjust the smoothness of the resulting function. It is really neat. I wrote some little programs to try this out and it works...
I visited several institutes in China, and had a talk titled "Connecting Data with Domain Knowledge". The slides can be found here
I had a very good discussion with David Tse, which lead to some new research topics.
I wrote a new wiki page on model selection and a geometric view of the Akaike Information Criterion.
I visited Virginia Tech, had a seminar and many very inspiring conversations.
I visited Lincoln Lab today, gave a talk on information theory and deep learning. Lincoln has a good group of people interested in the results and the applications.
- ‹ previous
- 3 of 3