2010-10-27
Revolutionary
2010-10-26
Chat Protocols and reliable delivery
I am slightly disturbed by software's inability to deal with frequent disconnections. Pidgin and other software should do a better job of resyncing after the fact. Anyway, my girlfriend and I have implemented failure detection overtop our conversations by manually numbering our messages so we can detect gaps now. Yay!
Horrifying Tales from the Academicrypt.
Over the last week, I
- did Lots Of Marking
- created a lab I didn't give (the students deserved better)
- studied and attended a midterm (which I arrived late for because I wasn't paying attention to my alarm!)
- gave a presentation on a case study on a paper
- worked tireless on a largish assignment for 3 days
- failed to submit the assignment
- did Lots More Marking and Remarking
I don't feel so good anymore, but that probably has more to do with my chaotic eating habits (I haven't had time to get groceries), and walking in the wet on Sunday for 2 hours carrying heavy baggage (Guelph buses don't run on Sunday evenings, and London's don't at 1:30AM). That's what I get for taking a train.
2010-10-25
New Perspectives
School, not Fur
2010-10-24
Mealing
2010-10-23
Clustering, Classification, Categorisation
I've divided my blog into 3 main categories that use tags to distinguish between them. There is now
#General contains my main posts, #Microblog contains what I formerly would have put on identi.ca (I plan on having it syndicated there) and #School is my attempt to keep track of my time (which all belongs to school anyway). I might create a #Photos too eventually.
I never liked writing really short blog posts, but I also dislike having my thoughts spread over the Internet in different places. I want to have One voice, which I might syndicate elsewhere. So, I'm letting blogger through Kosmokaryote be my interface to the world. Hello there.
I hope this leads to more postings here on my blog :) Sadly, I don't think I can yet create separate Atom/RSS feeds on a "category"/per-label basis. Sigh!
Brain Cuttings
2010-10-13
Comrades
2010-10-12
What's that in a name?
Progress?
It was Thanksgiving weekend, and that gave me time to finish my Natural Language Processing assignment. A simple scanner/lexer that required use of regular expressions, JFlex, and some thought. Parsed a simple .sgml file. The trickiest part was discovering the "/" look-ahead operator for handling apostrophized work. That said, having all this work would have been a greater hindrance to the holiday if it was not for my amazing girlfriend who did all the driving and even prepared none-feast meals for me as I sat in front of my laptop for hours and hours and hours at a table of food or in a moving car careening down the highway. I'm lucky :D
Left to do? Sort out this AI group assignment which is in trouble, I dare say, and try to manage to make some progress on my actual research :| Need to firm in stealing time for things.
2010-10-09
Fundamental Bugs
Another Busy Week
I was struggling to get Assignment 2 marked when it turned out I would need to hold a lab this past Thursday on networking, which I had expected/hoped to hold next week. The professor was great and offered to handle creating the assignment. (I sort of munged the last one, with too vague instructions on some points; I hope I get to create another one.)
The lab went off fairly well for my part. It was perhaps the best lab I've created yet, with the clearest example and clearest tutorial to accompany it. However, it probably went off worse overall as few people did their component, creating their own chat application. Sigh! I see the value in grading people on lab participation or lab quizzes :) Good luck to them on their next assignment?
Anyway, yesterday I went through and reviewed all the assignment 3 submissions, which is somewhat awkward, as I haven't gotten around to finishing assignment 2. (Work load has reached the saturation point.)
CIS*6650: Natural Language Processing
We have an assignment, which I have half-done but which is due tomorrow. I suppose that's my task for today. A classmate has been asking questions, sometimes largely in relation to Java, but which also cause me to think of new things on the assignment. Yay. I usually prefer to work independently, at least for the purity of my own thoughts, but I suppose new thoughts resulting from discussion can be useful :)
CIS*6020: Artificial Intelligence
This is beginning to feel like a small catastrophe :D I joined a group that intended on studying one thing, but then after discussion with the professor will probably now study something different and less interesting. Unfortunately, we can't seem to agree on what it was the professor was recommending, and our individual understandings directly contradict each other. Are we explicitly NOT measuring the performance of these characteristics, but rather using statistical measures (which?) to understanding why they make good features, or are we measuring the features' interdependence to determine their performance? It's been over a week and we're still not clear, and we have 5 days to prepare for a poster presentation. It gets worse in that our group seems to find it impossible to meet. I'm only on campus on Tuesdays and Thursdays, one group member was gone for the first week and the other one just went to Calgary. Yikes indeed. Wish us luck?
Research
Given the poster presentation, the NLP assignment, the lab preparation and assignment marking, I haven't had much time to make any progress on this, which is frustrating. I'd rather spend half of my total time on this each week, rather than near-zero.
General
My days on campus are almost always completely full, with any time in the long day between classes filled with meetings and administrative stuff (like OSAP and keys and forms). Also, all my other days are full too. I am intent on carving out something like a weekend for myself somehow. I think it might be easier next semester if I live near campus, so I can go and hide in a library on non-class weekdays? There isn't any nearby working spaces in London distinct from my leisurely home environment, not that I've had any time for leisure here. It might just mean I'd have to be more efficient with eating and cleaning in a day.
Happy Thanksgiving!
2010-10-01
Masters of the Universe
CIS*6020: Artificial Intelligence
Lectures on Tuesday and Thursday. Focussed around a group project. I initially tried to work with an acquaintance, but who ended up leaving the course. Then I tried to join a group with a linguistic focus, but it seems their interest has been swayed. Ah well. We have a poster presentation to prepare and perform in just over two weeks.
CIS*6650: Special Topics: Natural Language Processing
Fascinating. I even purchased the textbook. Lectures in the early morning of Thursday. We have an assignment coming up soon that involves ((j)f)lex.
Teaching assistanceship for CIS*3750: System Analysis and Design in Applications
In some ways, this is the most fun course. I conduct a lab once or nonce a week and have gotten to prepare an assignment so far, as well as comment on their first assignment. It's the most fun so far because it's had the most work, and the class's project is a fun one. They get to program an 18th century naval battle simulation/game.
I don't think I've been a great TA so far. I feel like my instructions have been too vague. I need to specify better, but I don't want to restrict their freedom. There have also been some hiccoughs. The first lab, I did not have the projector password. That didn't matter much, though, as the projector console was broken anyway. The second lab, we were the first group in a new computer lab in Reynolds, and there were hiccoughs with the students' accounts, like not having their home directories mounted.
They seem to be taking it well. We might have a mascot, Ozzy.
Oh, yes, my research!
Initially I felt a bit frustrated, as having arrived at Guelph, I didn't really know what was going on or what I needed to do for this. After the first couple of weeks, I finally got an idea, and have even made progress toward it.
We're looking at P53 (who isn't? :D) I am going to work on trying to determine which sites on the gene are most important to its function by comparing them to other genes which have different function. I think :)
Tools
Labels
Blog Archive
-
▼
2010
(70)
-
▼
October
(19)
- Revolutionary
- Chat Protocols and reliable delivery
- Horrifying Tales from the Academicrypt.
- New Perspectives
- School, not Fur
- Mealing
- V-V-Vonage
- Clustering, Classification, Categorisation
- Looxcie
- eNovella
- Brain Cuttings
- Eaters gonna eat. Voracious eaters gonna eat a lot.
- Comrades
- What's that in a name?
- Progress?
- Fundamental Bugs
- Another Busy Week
- Masters of the Universe
- Tools
-
▼
October
(19)