Sunday, November 06, 2011

b-matching as improvement of kNN

Below is an illustration of b-matching from (Huang,Jebara AISTATS 2007) paper. You start with a weighted graph and the goal is to connect each v to k u's to minimize total edge cost. If v's represent labelled datapoints, u's unlabeled and weights correspond to distances, this works as a robust version of kNN classifier (k=2 in the picture) because it prevents any datapoint from exhibiting too much influence.
They show that this restriction significantly improves robustness to changes in distribution between training and test set. See Figure 7 in that paper for an example with MNIST digits. This is just one of a series of intriguing papers on matchings that came out of Tony Jebara's lab, there's a nice overview on his page that ties them together.

10 comments:

Gokul Ravi said...

nice blog
android training in bangalore
ios training in bangalore
machine learning online training

Gokul Ravi said...

useful blog
python interview questions
cognos interview questions
perl interview questions
vlsi interview questions
web api interview questions
msbi interview questions

Gokul Ravi said...

laravel interview questions
aem interview questions
salesforce interview questions
oops abab interview questions
itil interview questions
informatica interview questions
extjs interview questions

Gokul Ravi said...

sap bi interview questions
hive interview questions
seo interview questions
as400 interview questions
wordpress interview questions
accounting interview questions
basic accounting and financial interview questions

Anonymous said...


nice blogs about financial accounting at The Basic Financial training in bangalore

Alamgir Hossain said...

Sometimes we are really wanted the improvement of our things but at the end of the period the things are like it was before. see more details to check out more helpful guide and tips for writing your papers.

Arthur Mendoza said...

I would like to know about the improvement of knn and those factors which is really matter for such an imporvement. helpful link that is very helpful for the academic papers writing.

Unknown said...

I cannot even believe that these kind of things would have been happened in the seller function of the college in here. http://www.qualitativeresearchcritique.com/our-critique-services/ to see more about the writing tips.

s suma said...

very good informative blog & useful to me thank you...keep posting
Machine Learning Training

Unknown said...

Better improvement we can do with this illustration and it will be good way to complete our task. I hope educators are also like this system and they can follow the http://www.doctoralresearchproposal.com/writing-a-dissertation-proposal/ for another guideline.