Category: Diary

It’s a shame that I haven’t written anything here for such a long time. I have done many things since my previous post. I learned many things, worked on a quite few subjects, went to the NIPS conference and attended a few workshops there (the first day was completely devoted to the New Trends in Reinforcement Learning. I sampled in the second day), went to the Ottawa in the holiday, and etc.
So, this post is just to say I’m still alive!
That’s it!

In this week, I am in Vancouver attending the World Congress on Computational Intelligence 2006 (WCCI). This congress consists of three important computational intelligence-oriented conferences: International Joint Conference on Neural Networks (IJCNN), Congress on Evolutionary Computation, and IEEE International Conference on Fuzzy Systems (FUZZ-IEEE).

Here, I try to write about the congress. I guess I cannot devote enough time to write very organized and well-written reports from the conference, but I try to note my impression about some aspects of it. This means that I try to give a pointer to the papers that I find interesting, write a few key sentences about the content of invited talks, and even mentioning that the coffee was good or not.

One important thing that I must emphasize is that I cannot attend all sessions for sure (and there are usually 15 parallel sessions). This means that I may miss many good papers. Anyway, let’s see what will happen.

-Have I evered told you how do I feel about you?!
+No fool.
-I hate you!

Last week, Pascal Poupart came to our university. He gave two talks and had several private meetings.
I wanted to talk to him sometime but unfortunately I could only attend one of his talks as the previous week was a hazard for me.
I was curious to talk to him about POMDP (which is his speciallity) and try to see the similarities/differences between POMDP and the concept of partial observablity in control theory from his perspective. For sure, both of them are talking about the same thing, but the goal of investigating PO is not completely the same. One reason might be due to discrete/continuous state differences between those two which leads to different approaches of the problem, i.e. state reconstruction might be easier if we assume that the dynamical system obeys some regularities such as linearity. Anyway …

Hi everybody after a long time!
It is almost a shame that I do not update here regularly. The worst thing is that I do not have a consistent plan about what I want to write here, what I want to tell in my Persian weblog, and what to keep just in my notebook. Well! Let’s not talk about administrative(!) stuff right here.

This semester is quite busy. I have taken the following courses:

Reinforcement Learning (Rich Sutton)
3D Vision (Martin Jagersand)
Operating System Concepts (Pawel Gburzynski)

Also, I audit Dale Schuurmans’ Machine Learning course. All these courses have their own readings, assignments, and so on! In addition, I want to do my own readings which certainly take much time. The other thing that has been taking my time recently is preparing a few papers for a conference and a journal. I do not want to mention all those seminars, meetings, and … that I should attend. Putting all these together result in being completely busy!

Doing a Ph.D., according to this funny articles, can strongly be easen by eating chocolate. I have not tested that in a systematic way (yes! I eat chocolate too!), but I’ll try to do so. If my weight increases in the near future, you will know what the reason is.

I will attend this NIPS 2005 conference this year. I try to log the conference. I don’t know whether I can do it or not. Anyway, let’s try it …

Today, suddenly I felt that I had a lot of work to do, and then became sad (or might be depressed)!!! I should do my numerical optimization assignment by Friday (and I did not started it then), prepare a poster for Friday’s 603 final poster session, have a discussion with Mohammad about feature selection in RL, write my technical report for the Robotics challenges course, do my course project for the numerical optimization course, some possible coding for the Robotics Challenges course, prepare slides for CDC-ECC, write rebuttal for my previous paper, proofread the Memetic/Evolutionary/Learning/… paper, go shopping, and read many papers and books that I want to read, and while enjoying my life as my previous professor told me when I was going to come there! Worst of all, this NIPS 2005 will start next week in the middle of all these jobs! However, I do not want to miss that conference. It is a good experience for me to visit my colleagues.

I have a very exciting assignment for my optimization course: calculate the Hessian matrix for a feedforward neural network. I don’t know if there is any simple trick for simplifying things, but the way I’m extending the usual backpropagation to the second order derivatives are really exhausting. I guess one might have published a paper just for this calculation a few years ago.
(At least, I hope that Bishop’s book has a solution for it. I really don’t want to implement a code without checking it before.)

Just now, I’m doing some experiments using gnuplot! It is a cool software IMO! Moreover, I tried Octave and it was cool too. It is somehow (or even more!) like MatLab with the same command lines – as far as I see. It is wonderful that these kinds of softwares are freely available (neglecting the fact that MatLab was also free for me too! 😀 ).
A very interesting phenomenon can be seen in my current life: I am facing a rush of new softwares, tools, and even culture! I guess this learning step is much bigger than whatever I’d in a few last years. Maybe not too! like this condition – at least now.