benfrederickson

Thanks!

If you really want to play around with it, all the code is on my github https://github.com/benfred/bens-blog-code/tree/master/distan... =)

I like matrix factorization models too !

Most of the netflix prize style models will have problems with this sort of data though. The problem here is that you only have positive data, and no negative data: you can assume that artists the user listened are positive, but there is no explicit negative votes for artists the user didn’t like. This means that there is no gradient for Funk style (http://sifter.org/~simon/journal/20061211.html) SGD solvers to descend.

Better matrix factorization models would be something like the weighted ALS (http://labs.yahoo.com/files/HuKorenVolinsky-ICDM08.pdf) or BPR (http://www.ismll.uni-hildesheim.de/pub/pdfs/Rendle_et_al2009...) approaches which can handle this sort of thing directly.

Having said all that - this post isn’t about CF. Its just trying to explain some basic concepts in information retrieval using finding similar artists as a problem.

Item-Item collaborative filtering could use these distance metrics when doing a nearest neighbours recommendation - but on its own not really CF. CF would provide results that are personalized to each user, while this just brings up a static list of results for everyone.

MSP is mandatory - but its tied to how much money you make. Most people here are probably making enough that they are paying the maximum amount (which is $72 for an individual these days), but as a university student I had the same coverage for free.

EI coverage last significantly longer than 15 weeks if you've been working full time in the last year. It depends on the regional unemployment rate - for BC its probably around 36 weeks: http://www.servicecanada.gc.ca/eng/ei/types/regular.shtml#lo... .

Software developer salaries are usually about 30% higher in the Bay area than they are in Vancouver.

Renting an apartment will cost you at least twice as much in San Francisco though - so most people are still financially better off with the smaller salary in Vancouver. There are non-financial reasons to move to the Bay Area (more interesting work being the big one), but don't expect your salary to go farther

Thanks for the feedback!

You're totally right that ratings datasets will skew positive. I was trying to get really get across that they aren't bimodal at all (unlike what some other people have claimed) and tend to look somewhat normal if you squint hard enough.

reddit has open sourced their code at https://github.com/reddit/reddit(https://github.com/reddit/r...

I wonder how this will play out will the publishers. As I understood it, they had a revenue sharing agreement with a large number of content producers: news.me charged money to their users, the publishers got a cut of it, and the users of news.me got to see the content in a clean 'readability' style view. Now that its free, I can't see how this will work out with their existing publisher agreements ...

This is ultimately the reason we chose Thrift over PB when developing our app (zite). Thrift was over 40 times faster in serialization for our python code.

Not to be pedantic, but the United States isn't bigger than Canada (its just more populous).

I use this library at my work, and I'm happy with it - but a word of warning: it's not trivial to integrate into your project. There is no makefile included here, and you'll have to modify the logging.cc/logging.h code to even get this to build under linux.