?

Log in

No account? Create an account

Catching up on another week, things are now moving on the MSc… - John C. Kirk — LiveJournal

Dec. 14th, 2003

06:30 pm

Previous Entry Share Next Entry

Catching up on another week, things are now moving on the MSc project.

Firstly, a small rant (a rantlet, if you will). The first step in this project is to select which project to do. In theory, the way this works is that the supervisors post a list (on Friday) of the projects they're offering. We then submit our choices by the end of January, and get assigned one (not necessarily our first choice, depending on how many people want to do it). So, I would like to approach this in a calm, reasoned, manner, by looking at all the available projects, discussing the ones that interest me with the relevant supervisors, and then submitting my choices after careful consideration. In practice, it's more like an Ebay auction, complete with sniping. The list of projects started appearing last week, and by Wednesday new ones were popping up every few hours. More to the point, some were also being booked within a couple of hours of appearing.

For instance, one project is to write a clone of MSN Messenger. This doesn't really have any relevance to my long term plans, but I think it's something that I could do well. So, I swapped emails with the supervisor on Wednesday evening and Thursday morning. At 10am on Thursday, he advised me to look at an open-source version of Messenger on the web, to see whether I could understand the code. Fair enough - I figured that I could start that in the evening when I got home from work. By 2pm that day, the project had been booked to someone else. Meh.

So, it's a balancing act of "do I pick this project now, or wait for a better one to come along?" The added complication is that there are projects like these, where the supervisor says "I will only discuss these on the 6th Jan". My compromise was to wait until Friday, so that I could get the full list, but I won't wait until January to hear presentations.

I've decided to put the semantic web stuff to one side for the moment. It's something that I may come back to later (e.g. for a PhD), but for now I think I'm better off picking a project that a supervisor has defined, rather than suggesting my own. That way, they'll have a better idea than I do of what's suitable, and I can hopefully get experience of a successful project, which I can use to come up with my own topic later.

Anyway, I'm currently leaning towards an Artificial Life project - one of the first two on this page. I bought a copy of "The Blind Watchmaker" on Friday (I've previously read "The Selfish Gene", also by Richard Dawkins), so I'll start reading that later today. One slight concern I have is that I'm not a biologist, so I know very little about hyenas etc. So, I may be better off going for non-existent creatures such as the biomorphs. On the other hand, the ant colony idea could be interesting. I might also talk to the supervisor about modelling bees, having read Clan Apis a while back, since that would have similar themes, such as pheremones.

I'm also a bit more comfortable thinking of insects as mindless automatons, whereas I like to think of animals as sentient. Mind you, there was an interesting idea in "Godel, Escher, and Bach" (by Douglas Hofstadter), where he suggested that an ant-hill could be sentient, even though the individual ants are not (they are the equivalent of neurons in the human brain). The concept here is emergent intelligence. I really ought to get round to finishing that book some time, since I bought it about 10 years ago, and the furthest I've got is half-way through...

One thing I remember from the project book I read a while back is the idea that there shouldn't be any bad results. In other words, rather than relying on one particular outcome, I should plan it so that any result is useful (even if it is "in this scenario, everyone winds up dead"). So, I'll try to bear that in mind.

Anyway, on the plus side, I've felt a lot more energised over the last few days, since project stuff started happening. I'm taking a week off work over Christmas/New Year, so assuming that I've picked a project by then, this should give me a good opportunity to start by background reading.

On a side note, I also liked Jay Hosler's science cartoons, e.g. "Darwin saves the world!" And Killer Bee is effectively a trailer for "Clan Apis".

On the social front, I went off to a Christmas party in Putney on Wednesday, which was a pleasant evening. Then we had a gathering here yesterday, which went well.

This afternoon, I went off to watch Freaky Friday at the cinema. I felt slightly embarrassed about being the only adult there without an accompanying child, but I enjoyed the film. It's not intended to be one of the great all-time classics, but I think it achieves what it sets out to do; it's a fun way to spend a couple of hours. The cast do a good job, and I particularly liked the well-timed rugby tackles.

Comments:

[User Picture]
From:rileen
Date:December 15th, 2003 06:29 am (UTC)
(Link)
Hi, regarding the projects, i'd say

(1)The MSN msgr thingy doesn't seem like an M.Sc project anyway.

(2)Good that you won't work with someone as fickle as that !

(3)The '6th Jan guy' seems rather inflexible & perhaps not the most approachable.

What about Colin?
(Reply) (Thread)
[User Picture]
From:johnckirk
Date:December 16th, 2003 02:34 pm (UTC)
(Link)
(1)The MSN msgr thingy doesn't seem like an M.Sc project anyway.

I know what you mean - I was quite surprised when I saw it. On the other hand, if I see a project that's a complete gift, it would be silly to turn it down...

(2)Good that you won't work with someone as fickle as that !

Well, I don't think he's entirely unique. Something similar happened with Mary Dalrymple's projects. In her case, she emailed me back with more details, and I decided that I didn't like them, so it wasn't a major loss, but there's still rapid movement going on.

My only concern about Richard Overill as a supervisor is that he was in charge of the networking module last year. He seemed like a good lecturer, from the initial lecture that I went to. However, I know that a lot of people failed that exam, who did well elsewhere, so I suspect that it may have been similar to the DSI course, where the lectures didn't really prepare them for the exam.

(3)The '6th Jan guy' seems rather inflexible & perhaps not the most approachable.

I know what you mean. Mind you, Maxime won't see people until the 23rd of Jan! (Or the "23th of Jan" according to his page.)

What about Colin?

I did email him, after Yoan said that I might be able to get someone else to supervise the Messenger clone project. That seemed odd (I thought all projects were supposed to be unique), but he (Yoan) has assigned one of his other ones to two different people, and I guess that if you're writing a clone then it's not going to be that unique anyway. Colin said that he would only supervise the projects he's offered, as he doesn't have time to learn any new stuff. That's reasonable, but none of his projects particularly appeal to me. And given that I got my lowest mark in Random Algorithms, I think I'd do better in a different area.
(Reply) (Parent) (Thread)
[User Picture]
From:rileen
Date:December 17th, 2003 06:42 am (UTC)

Some gifts should be turned down !

(Link)
On the other hand, if I see a project that's a complete gift, it would be silly to turn it down...

Would it ? What can it do for you, besides possible fetching you good marks for less work - i fact, it may be a lot of boring work. Moreover, if you still intend to do AI research, i doubt this fill enhance your prospects. No, a "gift" of this kind seems to have little value in the long run.

It's amazing how so many of the supervisors at KCL seem so disinterested & far from approachable - i think the department would do much better with healthier attitudes.

Maxime is rarely around - do think about that. On the other hand, Sergio got excellent marks under him, so it's an interesting thing to think about, whether to work with Maxime.
(Reply) (Parent) (Thread)
[User Picture]
From:rileen
Date:December 17th, 2003 06:48 am (UTC)
(Link)
Sorry about the typos - 'fill' instead of 'will', 'possible' instead of 'possibly' .....

At least i have the excellent excuse of having watched the LOTR triple till 3:30 am :-)
(Reply) (Parent) (Thread)
[User Picture]
From:johnckirk
Date:December 17th, 2003 04:52 pm (UTC)
(Link)
At least i have the excellent excuse of having watched the LOTR triple till 3:30 am :-)

Hope you enjoyed it - I won't be seeing the third film until the weekend, but I'll try to have my own marathon by watching the DVDs earlier in the day.
(Reply) (Parent) (Thread)
[User Picture]
From:rileen
Date:December 18th, 2003 01:25 am (UTC)
(Link)
Oh, i sure did - and perhaps your marathon might be longer, if the DVD versions are extended ones?
(Reply) (Parent) (Thread)
[User Picture]
From:johnckirk
Date:December 17th, 2003 04:57 pm (UTC)

Re: Some gifts should be turned down !

(Link)
Well, first up, this is now a moot point, since I've booked the ecological artificial life project (I'll post more about that in a new entry). But thanks for the advice anyway.

But from a theoretical point of view, it's hard to say what the long term value would be. E.g. I thought that some of the questions in the DSI exam last year were complete gifts (since they were pretty similar to the stuff I'd been doing at work for so many years), which probably explains why I got such a good grade. However, people who look at my grades later won't know (or care) what the questions were, they'll only look at the actual result. The rules are slightly different for projects, but I think it depends whether I wind up doing a PhD, or whether I carry on in commercial programming. If I'm just looking at it from a CV point of view, then saying "I did an MSc, and got 90% in my project" would justify the time off work, and I don't think they'd enquire further.
(Reply) (Parent) (Thread)
[User Picture]
From:rileen
Date:December 18th, 2003 01:28 am (UTC)
(Link)
You're right, for a job, the marks are all that should matter, though there might be exceptions. Good marks in the project can be used as an indication of good programming skills, and other positives.

The title of your project sounds interesting - look forward to reading about it in your post(s).
(Reply) (Parent) (Thread)