Skip to yearly menu bar Skip to main content


Talk
in
Workshop: Nearest Neighbors for Modern Applications with Massive Data: An Age-old Solution with New Challenges

How to stop worrying and learn to love Nearest Neighbors

Alexei Efros


Abstract:

Nearest neighbors is an algorithm everyone loves to hate.   It's too trivial, too brute-force, doesn't offer any insights.  In this talk, I will try to make you fall in love with the humble nearest neighbor.  First, I will discuss the use of nearest neighbor as an exceptionally useful baseline to guard against our field's natural bias in favor of elegant algorithms over data.  Then, I will discuss some scenarios when nearest neighbors is particularly useful in practice.

Live content is unavailable. Log in and register to view live content