Rate-distortion, Wyner-Ziv, typicality, low-density codes, etc.
Information theory is, in my opinion, basically the study of Euclidean spaces with a huge dimensionality, particularly problems like sphere covering and sphere packing. Many things are easier when the dimensionality approaches infinity, but unfortunately such a space is quite hard to visualize and usually impossible to deal with directly using a computer (how would you calculate the Voronoi regions of 2^1000 points?), so it can be intimidatingly abstract and theoretical at first. However, once one learns to visualize such spaces, the theory becomes quite delightful to study.
* * *
I'm having some difficulty accessing many U.S. sites (MSN in particular) these days, because the Taiwan earthquake broke many fiber-optic cables. Don't feel surprised if there is no update for one week or more.
* * *
And happy birthday, H. You are the dearest friend one could ever have.
* * *
I'm having some difficulty accessing many U.S. sites (MSN in particular) these days, because the Taiwan earthquake broke many fiber-optic cables. Don't feel surprised if there is no update for one week or more.
* * *
And happy birthday, H. You are the dearest friend one could ever have.