Getting back to research after break

The winter quarter started a week ago. I have my classes pretty much figured out at this point; now I am trying to catch up on my research from where I left off before break. As I am ruffling through my messy notes, I realize going on break suddenly made all my research projects seem very distant. But, on a good note, I am getting back on track again.

Happy New Year!

A few more hours to 2010. I wish everyone a great new year ahead; hope all your dreams, aspirations, and resolutions are fulfilled. Hope the world becomes a more better place in 2010.

Happy New Year!

Broken internet and The Mystic Masseur

Going back to reading books over break has been refreshing. I just finished reading Naipaul's The Mystic Masseur. I don't think I ever read a book by Naipaul before, but I have to say that I am pretty impressed. I like the imagery that permeates Naipaul's writing. The Internet in my apartment was down for the last few days, and that definitely put me in the mindset to pick up a book. I also realized that my life has been deeply interwined with the Internet; I felt something was amiss in my life, that some portion of my daily life has suddenly disappeared - an appalling dependency on technology.

The net is back, since I am writing this entry, but the irony is that I need the technology to talk about my appalling dependency on the technology.

Lifted first order and first order languages

I spent a week or two exploring first order and lifted first order languages. The idea is very simple; express models in first order, and the system creates the necessary graphical models to perform inference. From a machine learning perspective, this is a very useful idea. Since graphical models are not always in the repertoire of machine learners, this kind of first order language specification of probabilistic models will be advantageous. I think the eventual goal is to have such a system as a library to a standard programming platform, where the complexity of graphical models is abstracted away in the form of a blackbox.

AI taking over?

Here's something that is worth reading.

Transfer learning and kernel based multitask learning

I have a number of papers on transfer learning that I have started looking through for one of my research projects. Transfer learning is a machine learning technique where the basic idea is to learn multiple related tasks simultaneously, rather than learn each task independently. Previous work have shown that this works better, especially for predictive performance, compared to independent, single task learning. For my problem, I am interested in using multitask learning techniques based on natural extensions of kernel based methods for single task learning, such as the Support Vector Machines (SVMs).

Go programming language

One of my friends from the compiler group sent me an email regarding Google's new programming language, Go. From the Google talk presentation it seems this language has some unique and amazing features, and most interestingly, it feels like a language where one can code like python and get power of C++. Check out this talk if you want to know more about the language.

Procrastination weeks

Procrastination has been an old friend of mine, going back a long way. We lost touch at some point in college, except occasionally crossing each others path. The last two weeks of this quarter, procrastination has become my constant new friend. It's probably time to cut loose of this friend, except the occasional visits.

Why $\textstyle l_2$ norm?

Recent progress in signal processing theory has generated a renewed interest in the effectiveness of using the $\textstyle l_1$ norm. It is interesting to note that, for example, in sparse signal recovery one can do a more accurate signal reconstruction minimizing the $\textstyle l_1$ norm.

The reason I have been thinking about $\textstyle l_1$ norm is because it apparently looks more useful for a portion of my current research. Maybe over a hundred years ago, or so, whenever mathematicians started thinking about norms $\textstyle l_2$ won, but current research trends show that $\textstyle l_1$ norm is very handy for a number of practical purposes.

Convex optimization reading group

I am now part of a reading group on convex optimization. We are going through the amazing optimization book by Professor Boyd.