What is Digg?24 Comments
- Sonic_Molson, on 10/12/2007, -8/+25No offense to Kevin, but I don't see why EVERY article posted by him has to end up on the front page.
Kevin, submit under a different username sometime, and then reveal the results on diggnation. - PlanarPurple, on 10/12/2007, -2/+12Although the authors have a good overview of clustering, I could not find open-source implementations. I have applied Kolmogorov Complexity to data compression for hierarchical clustering in the scientific papers listed at
http://cilibrar.com/
and have an open-source artificial intelligence library package called CompLearn available from
http://complearn.org/
CompLearn lets you do hierarchical clustering and other forms of machine learning based on data compression and quartet tree search heuristics. It has been used to do fully automatic evolutionary tree construction, musical recognition, language analysis, and many more fields. There is even a drag-and-drop demo and online demo that you can try right now. Surprisingly, the same techniques can be attached to Google to provide subjective analysis of search terms using hierarchical clustering. - SamuelClemens, on 10/12/2007, -0/+4A good intro on Self Organizing maps.
http://www.ai-junkie.com/ann/som/som1.html - MrUnderbridge, on 10/12/2007, -2/+5Check out Kohonen's Self Organizing Map (SOM) algorithms, at
http://www.cis.hut.fi/projects/somtoolbox/
This is the matlab version of his toolbox, but there's a C version around as well. GPL licensed. In fact, Google "GPL" and "Clustering" and you'll find rather a lot. - mapkinase, on 10/12/2007, -2/+5That is lame. I am in clustering and you can find better sites just by googling.
- inactive, on 10/12/2007, -0/+3@1021,
Could you provide us with a better example? I have a relatively new interest in the topic. - sentanta, on 10/12/2007, -0/+2YALE (Yet Another Learning Environmnet) is good too -
- nofxjunkee, on 10/12/2007, -0/+2not to mention that there are clearly applications of clustering that apply to digg...
- mandel15, on 10/12/2007, -0/+1It is not a PhD paper in clustering but it does give some introduction to the theme. Lets face it, maybe some wants to get in the clustering world and has no idea, this is a good enough page to give him an intro. It definitely does not say anything advanced, but common who goes to digg to find scientific papers,you use google scholar or something like that instead.
- Snakedal337, on 10/12/2007, -6/+7Because hes got a huge ass network. As in, nearly every person on digg.
- erikuma, on 10/12/2007, -0/+1Nice resources! Thanks! Especially the cilibrar.com.
- 1021, on 10/12/2007, -4/+5I would say this is one of the worst-done descriptions I have seen about clustering. Really bad job googling by Rose on this one.
- Nizz, on 10/12/2007, -0/+1What an awful link! I did my PhD on Clustering and all I can tell is that this kind of tutorials are lame! Just by Googling any result would be better than what Kevin pointed to!
Check the bunch of Moore tutorials about machine learning, data mining: I used them during my state of the art phase and were really valuable:
http://www.autonlab.org/tutorials/ (this is about data mining in general)
http://www.autonlab.org/tutorials/kmeans.html (k-means and hierarchical clustering)
Online lectures on clustering, classification:
1-Dealing with Data: Classification, Clustering and Ranking
http://www.researchchannel.org/prog/displayevent.aspx?rID=4840&fID=345
A new trend in clustering called semi-supervised clustering:
Here's an online lecture I recommend:
1-Semi-Supervised Clustering: Probabilistic Models, Algorithms and Experiments
http://www.researchchannel.org/prog/displayevent.aspx?rID=4423&fID=345
Still guys I have a ton of links which I collected during 4 years working on clustering algorithms so anyone interested feel free to ask me!
ps: Sorry Kevin but your link offended me!!! - Itkovian, on 10/12/2007, -2/+2You might want to check out Weka, R, etc.
- amikael9999, on 10/12/2007, -0/+0Very interesting one!!!
Here are free books on Algorithm:
http://freecomputerbooks.com/compscAlgorithmBooksIndex.html - kordless, on 10/12/2007, -1/+1It is a non-trivial problem to take a mathematical notation and convert it to an algorithm for use in a code base. Kevin is interested in clustering, so he probably found this and though he'd share it with us.
I don't know much, but I bet he didn't take the algorithm presented and convert it to code to use in digg. The problem is that neither can we, and we are all looking for that instant gratification fix that only Digg can give us! ;)
So, in that vein, here's a cool little ap that you can run on Windows - all you need is a videocam: http://www.roborealm.com/ Realtime "clustering" of objects that your computer can see! - CharlesDarwin, on 10/12/2007, -2/+2But it's still better than his 'I'm blogging' video. ;)
- nofxjunkee, on 10/12/2007, -2/+2Who cares if everything he posts gets to the front page? I can spot 4 articles near this one that are 10x less interesting. If you like, digg. If not then just move on man.
- CDWMobile, on 10/12/2007, -0/+0I love being the 5th thumbs down - Spend more time on ArcGIS and less on Digg ;)
- lewhich, on 10/12/2007, -0/+0This is a nice article on Data Mining (knowledge discovery & AI)
http://dataml.net/datamining/web/toc.htm - misterpony, on 10/12/2007, -4/+1digg algorithm = overlapping clustering algorithm * probabilistic clustering algorithm * real-time sorting algorithm
/talking out my ass - saikhan, on 10/12/2007, -7/+3This article is a cluster of *****.
Yes, I know who he is you fanboy ***** suckers. - Kinneas12, on 10/12/2007, -6/+1THIS IS SO WEIRD. I've just completed a remote sensing course for my major and this was a very large part of it. I NEVER thought something like this would actually be digg worthy. It's sad that they don't talk enough about how much of a necessity this is for satellite imagery classification. This is why remote sensing using satellite imagery is so useful! :P
- cookiebearo, on 10/12/2007, -13/+1wooooot
kevin rose, i put you in my top 8 on myspace
+digg
Browsing Digg on your phone just got easier with our enhancements to the 
© Digg Inc. 2009
— Content created and posted by Digg users is