First steps towards solid research

Alright friends, I am blogging my first steps towards some meaningful research
today. :))))))

I walked into the lab today and Vineet tells me that our classifier is doing a
100% accurate classification feeding off itself. Yes, I know that does not mean
much, but this opens a whole new set of possibilities. (For those not into data
mining, take my word for it, for the experts, just bear !!)

Some of the things we plan to look at
1. Can we boost the accuracy with Intelligent Keyword selection
2. Can we actually use incremental learning algorithms to induce decision trees
3. Do we need to bias the filters in filtering out junk email towards more
conservative filtering?

Now that we have the infrastructure in place we can answer all these questions.
Thanks Vineet for getting us this far!!!

Discriminative Classifiers

In an attempt to model the brain, Computer Scientists came up with the Neural
Network. On of the interesting things I learn about an NN is that it takes 15
days to ‘learn’ one task or problem (blunt but a laymans point of view).
So that would make it probably comparable to the learning rate of a child …
are we there yet?

Algorithms in Computer Science

The algorithms exam really had me thinking for some time. Although I lost most
marks to carelessness, I really did not have an idea to solve this problem "
Give an O (V+E) algorithm to find shortest paths from a single source for
a directed weighted graph with edge weights only 1 and 2." As it ironically
turns out, this brainteaser has a sarcastically easy solution. If you pester me
enough I shall reveal it to you 🙂

This weekend is a busy one, with work required on the project and on a proposal
for research studies.

Theory Of Computer Algorithms

I was working on solving an algorithm for detecting odd length simple cycles in a digraph. Well made some headway, some other people did come up with a perfect solution though. Requires a lot of creativity and focus. Two friends of mine went down for plagiarism yesterday. Advice to Indian students coming to the USA for education. Do your own work man, your paying so much money to learn, atleast try and make a goddamn effort. Anyway, Indian students are now officially all cheaters. How unfortunate. Nevermind all that, I have some ideas for a Distributed AI project and I will make a proposal today.

Homework

I am swamped with it for now. So cannot blog much, hopefully will come back
to my slightly creative writing soon. Until then, I must apply my creativity on
Theory of Computer Algorithms, in determining if cycles of odd-length exist
in DiGraphs.

I have started on my Data Mining project.

Wednesday Morning

Things are going along really easy this week. But I still have to make a decision about which courses I am going to take this quarter. The dilemma was to pick between DCN-II and DAI. The decision just got harder to make when I attended the Intro to DAI lecture on Tuesday. Michael demonstrated the use of DAI in the making of Lord Of the Rings. The DCN-II Home Page is here!

Data Mining and Theory of Computer Algorithms cannot be dropped as I really feel they are really important and should be taken this quarter itself.

The weather in Rochester really really is downright extreme. Going to have to dress warm today.

Kudos !!

My work with STARE received an initial acknowledgement today 🙂 from
Prof. Ankur Teredesai ! Way to go.

I finally have a clear specifications for Project 3. The draft for the project will be revised and posted here.

My Distributed Systems Professor Prof. Alan Kaminsky is off to Seattle to attend OOPSLA and present a paper on Many2Many Invocation.

Today out of the blue, as I unwound in the library, I remembered myself at the age 14 (’92). On a fine summer afternoon. Mom was busy in the kitchen. I was hacking away on my first PC game on my XT. At that point I wanted to move a spaceship up/down/left/right so as to blast the incoming alien ships. Two problems dogged me, how do I use the cursor keys and how I might be able to ‘explode’ the alien spacecraft when they ‘contact’ with my lethal laser blasts. It took me some time to figure both out, I did it in both QBasic and in C. The point is not the solution to the problem, but the happiness that wrought my mind 🙂 at that time. I felt as if I had put not just 2 and 2
together, but 2 raised to 128 and 3 raised to 256 together. hehehe.
The spaceship would move and throw laser bolts at alien ships which would then explode. Jeez it was hilarious, I had to tell the first person I met, my mom.

In my opinion such happiness is true and is unfortunately now elusive. As an adult, satisfaction in my work is now masked and mired by deadlines and other peripheral matters, I don’t believe I can find the same thrill in my work that often. I desire to treasure that moment for the rest of my life. In the future I hope to find the same pleasure in my work.

Related Stories (2nd Feb. 2006): “Re-Igniting the Passion” – Creating Passionate Users. Kathy argues that the passion I find elusive is really inherent and that one cannot blame their job or their environment for letting that passion and enthusiasm slide. It is really yours to hold on to in the myriad of jobs that you will yourself indicate as part of your everyday routine.

Halloween

Today is Halloween ! Everyones dressed up in neat costumes. I think I even
noticed Cleopatra on Capmus. Jasons there too, so are the Pirates. Yesterday,
I turned in my paper and I should say it is still in draft stages.
I think I aimed too high and the paper turned out to be quite a mess of ideas.
I am worried about the grade I might get on the paper. I hope it will be fine.

I have a lot to do this weekend (again !!!).

I will be investigating the use of Bayesian Networks to filter spam.
Paul Graham has an excellent article on the same topic.
Good to see that Microsoft research is being rivalled in the Open Source
community.
Paul Grahams work has also been implemented.

The STARE web page needs a lot of work too.

On Monday I maybe initiated as a proper research assistant and STARE might
get a lot of exposure with the higher levels in RIT. I also have to build a specification for
Distributed Systems Project # 3 and on similar lines a client plugin for Outlook(TOP SECRET).

It was a tough week thus, I just about managed to submit the paper. Two days
of not having slept is now catching up with me, I do feel really washed out
and relieved that the weekend is near. I need to Plan more meticulously.
Things probably won’t fall apart so easily then. Maybe I get the feeling that
I am overcommitting myself.

More Links:
Spam filtering Resources at MIT