Powered by glolg
Display Preferences Most Recent Entries Chatterbox Blog Links Site Statistics Category Tags About Me, Myself and Gilbert XML RSS Feed
Thursday, Aug 21, 2014 - 23:26 SGT
Posted By: Gilbert

Singa Sera Que

So, I'm back. It was amazing how alien my desktop monitor appeared after several weeks of staring at a ten-inch netbook screen, not to mention how the runny nose has returned after its complete overseas absence (I'm suspecting the humidity), and how my body clock's gone from getting used to waking at 8-plus without an alarm, to getting drowsy at seven p.m. before waking fresh as a daisy at two; this caused me to miss United's season opener against Swansea on Saturday (which they lost anyway - as a small consolation, Ashley Young was determined to have dodged an aerial bombardment, contrary to his fans' worst fears)

It goes without saying that there's a lot of material to cover - 128 (fairly sparse) pages of notes from the trip, to begin with - so catching up may take awhile. After some deliberation, I've settled on separating the technical details and associated ruminations from the more general recollections as far as is possible, since the former may not mesh well with a chronological retelling (and, let's face it, is likely to cause the eyes of a large section of the audience to glaze over)

As a (rather appropriate) prelude to the trip, I had volunteered for the Big Data and Analytics in Healthcare conference at Marina Bay Sands (fun trivia fact: it's by some distance the most expensive building in the world; supposedly, the Burj Khalifa, despite being the tallest skyscraper in the world, cost barely a quarter of MBS - talk about high land costs!), in large part to sit in on the talks.

All considered, I probably did less than I was fully comfortable with, but I suppose the organizers were safely overmanned; still, I did get some practice at ushering, where I sought to cultivate that fine sense of discretion, or the balance between subtle acknowledgement for guests who knew (or preferred to find) their way, and helpful approachability for those that didn't, while distinguishing those attending the neighbouring Adobe conference instead.

Other than that, there was some mingling with a bunch of pharmacy students (the volunteers were a fairly diverse bunch), with a slight boo-boo when I claimed not to have a thumbdrive upon urgent request, without realising that I hadn't cleared my bag out yet. Oops. Soon found that the Nvidia Tesla GPU promoters from last year's HPC in Finance event had a stall here too, surely worth some marks for effort.


Canada Bound


Can you imagine Asian airlines coming out with videos like this?
[N.B. It might have been the first I actually bothered to finish watching]


After that, it was time for my first visit to the New World (as well as my parents', who insisted on tagging along). Mindful of the toll that four legs (Singapore-Narita-Detroit-Toronto-Quebec City) could take, I slept in for the first, before not being able to resist the in-flight entertainment; I'd wanted to watch The Monuments Men ever since that single golden exchange in the trailer, while Kick Ass 2 and The Great Gatsby could be considered some New York 101. Uh, was that WWE's Paige in 300: Rise of an Empire (no)? And eh, are the A-Team bloody flying a main battle tank by firing off shells (yes)?

At that instant, I felt an undefinably deep sense of guilt for complaining that a squad leader leaping from one rooftop to another in Ruse of Engagement was unrealistic.

Anyway, we arrived in Quebec without a hitch and... um, wait, where's our baggage? As fate would have it, a lady attending CNS 2014 was in the same straits, and we eventually left for our lodgings at Laval University. A short wait and chat with a Wisconsin-Green Bay visitor meeting her Israeli friend at the conference later, I retired to my room for the stay, and was pleasantly greeted by an envelope containing breakfast chits slid under the door the next morning - one set per day!


I COLLECT THEM ALL


Quite fortunately, I had seen fit to arrive a day early, which allowed the opportunity to shop for replacement clothes at Laurier Quebec (supposedly one of eastern Canada's largest malls) with the missing luggage entitlement. The Californian cashier at the university grocery shop impressed with his fluency in both Chinese and Bahasa Melayu, after which the scouting-out of the convention centre was complicated by rain. The airline got the stuff back to us that day anyway.

The big day having come, I made it a point to arrive for the workshop ahead of time (and claim a power outlet). A bit of an adventure with the projector later, the room slowly began to fill, and the fellow who took the next seat demonstrated his experience by furnishing an extension cord (apparently, the best way to make friends at conferences; it was only rather later that I realised that he had been a program cochair when I was still snivelling away in primary school)

I won't delve into the content, as explained, other than to note the Churchillian allusion in the naming of an depression-detecting phone app that was presented (frankly, they've become scarily predictive - my Google Now cards began supplying me with the Yankees' schedule, after at most a cursory search before I reserved a ticket, and the Metro Bus I had to take to the convention centre, out of nowhere!); having said that, the concluding panel discussion was extremely insightful, with some very helpful tips on navigating being an interdisciplinary researcher caught in no man's land (or: "what is this medical computing s**t all about?!")

Basically, the issue is that if one targets healthcare journals while primarily being a computer science person, there is often the issue of having to meet extremely exacting clinical validity tests, while missing out on one's primary audience (and, perhaps more importantly, one's tenure committee). On the other hand, the work may not get into CS venues due to insufficient novelty and generalizability, leaving aside the mirrored loss of credit for collaborating medical doctors.

One of the panellists gave the very relatable example of a former student, who achieved some exciting results on a cancer dataset, but with relatively simple and known computer science techniques. Of course, one might think that this shouldn't be a problem - should a new drug be discounted because its creators didn't discover a new extension to chemical synthesis along with it? - but as it turns out, it does if one wants to advance an academic career.

Things may be looking up, though, as more medical practitioners begin to recognize the potential of mining the avalanche of data they're amassing, and it was advised to always be on the lookout for networking opportunities and possible champions on the physicians' side, possibly by first fulfilling some tangible if not that super-exciting need.

The next day was spent on the From Deep Blue to Monte Carlo tutorial, the introduction of which was probably familiar to most who have taken an AI module, but I guess that can't be helped. Siri keynote in the evening, picked up some hamster feed from a pet store with a giant dog during the lunch hour, and some lumbering around the sea of posters in between:


Socially awkward penguin makes an awkward cameo


Split my time roughly evenly between the machine learning and game theory sessions (with a bit of vision) for the remaining days - among the major AI conferences, I suspect that the overlap with economics there should be among the greatest; one of the Senior Member presentations was on the implications of high-frequency trading, in which the speaker gamely gave voice to his critics as well. More next time.

There were also some refreshingly general big-picture sessions, such as the What's Hot in Cognitive Science talk that revisited the classic trolley dilemma, the engaging hands-on Behavioural Network Science demonstration, and how tablets are encouraging education in Africa, among others. Even better, I found some Malaysian noodles.

I was debating whether to give the speed dating (i.e. networking) initiative a go, and finally thought, heck, why not? The organizers had set up a series of long tables, at which participants were expected to sort ourselves according to our Chinese zodiac (helpfully posted at the entrance)

Started off with a Canadian fourth-year student, with whom I tried some of the suggested prompts (sadly, he didn't like Perl); next up was an Aussie, who was doing work on suspense. On hindsight, there probably was a better way to admit that one hadn't heard of this line of research, even if I meant it in a good way. A rather senior European prof followed, who mentioned some pioneer I wasn't aware of either. Oops. He did get the stress of developing medical solutions (and possibly misdiagnosing) spot on.

Fourth was with an USC guy who was doing something with Oculus Rift, and then a three-way with a prof from New Hampshire, who asked about my plans for after graduation (hmm), and finally an SMU student (small, small world)



comments (0) - email - share - print - direct link
trackbacks (3) - trackback url


Next: The Three Amigos: Amigo One


Related Posts:
On Economics
Lots Of Postlets
Conference Report
Davydov And Pratchett
Take A Seat Long Read
Back to top




3 trackbacks

Trackback by Sacs Longchamp Arbre de vie
Sacs Longchamp Arbre de vie - [bert's blog]


August 27, 2014 - 01:20 SGT     
Trackback by google plus api
google plus api - [bert's blog]


September 11, 2014 - 04:43 SGT     
Trackback by Forge Of Empires Trucchi
Forge Of Empires Trucchi - [bert's blog]


September 21, 2014 - 04:04 SGT     

Copyright © 2006-2025 GLYS. All Rights Reserved.