|
A rooms browser and selection system for my college (basically halls of residence). Front page  Example staircase of rooms  Example room, with survey from individual currently in there  When it comes to your turn to choose a room, you get this page  We ran it over the weekend for people choosing rooms for their third year and it performed very well - though maybe it was a tad heavy on the server with people repeatedly refreshing to see who's chosen what room.
|
#
¿
May 7, 2008 21:30
|
|
|
|
| # ¿ Apr 28, 2024 13:54 |
|
|
This is a probably bit more abstract and less flashy than some stuff posted here but still interesting. I'm currently implementing the Ising model of ferromagnets for a short uni computational project. The graph below shows how the final equilibrium energy of the system (bottom lines) varies with temperature, and the upper lines show the differential of the bottom line, effectively the heat capacity (how much energy you need to put in to change the temperature) Marked with a vertical line is the theoretical critical temperature above which the ferromagnet doesn't spontaneously magnetise. As you can see, the simulation seems to agree with the analytic prediction, which is good! The actual code is a neat little templated class that can handle any number of dimensions of grid and different wrapping systems. I've had a play with making it use OpenMP too (though the number of iterations isn't really worth it for the startup time, having done speed comparisons).  edit: Forgot to note that the yellow line is the variance of the energy at equilibrium divided by the temperature squared, which should be the same as the heat capacity according to the fluctuation-dissipation theorem (well, according to the project summary) - and turns out to be for this model! (which is nice.) vvv 'Cuz there's something wrong with my smoothing utility, initial conditions issue I think, need to fix that. chips fucked around with this message at 01:56 on Mar 25, 2009 |
|
#
¿
Mar 25, 2009 00:24
|
|
|
Sensible Thursday posted:To speed up the vocabulary-learning process, I'm finishing up a Chrome extension that uses the Google Translate API to switch random words on all the web pages I visit into an arbitrary foreign language. The theory is that sooner or later I'll pick up their meaning from context. Doesn't take care of the whole grammar problem, but it's better than nothing. This is an amazing idea - I love the idea of learning vocab in the context of your own language. Might not be so effective for grammar, but still could be useful. I suck at school taught languages, but I might just have to try this.
|
|
#
¿
Feb 16, 2010 12:25
|
|
|
Contains Acetone posted:Well life got in the way. Code has now been submitted, and I've confirmed I can actually pull it down and build Release and Debug. Come check it out! This is really cool -- I've been learning about ML recently and am keen to have my first go at a RBM. Any advice on good starting training sets?
|
|
#
¿
Sep 3, 2012 23:50
|
|