This preview shows page 1. Sign up to view the full content.
Unformatted text preview: .3: Circles and ellipses p.235 Sometimes it’s useful to use different scales for different
8. Distance-based models Given an instance space X , a distance metric is a function Dis : X £ X ! R
such that for any x , y , z 2 X : Therefore: use an ellipse rather than a circle to identify points p.235 t distances between a point and itself are zero: Dis(x , x ) = 0; t all other distances are larger than zero: if x 6= y then Dis(x , y ) > 0; Also consider rotating the ellipse.
p t distances are symmetric: Dis( y , x ) = Dis(x , y ); t detours can not shorten the distance: Dis(x , z ) ∑ Dis(x , y ) + Dis( y , z ). August 25, 2012 p 2
2 ! 45 degree clockwise rotation
connecting points at order-p Minkowski distance 1 from the origin for (from
inside) p = 0.8; p = 1 (Manhattan distance, the rotated square in red); p = 1.5; p = 2 S= 01
(Euclidean distance, the violet circle); p = 4; p = 8; and p = 1 (Chebyshev distance, the The Minkowski distance with p<1 does not satisfy the triangle
Machine Learning: Making Sense of D...
View Full Document
This note was uploaded on 02/10/2014 for the course CS 545 taught by Professor Anderson,c during the Fall '08 term at Colorado State.
- Fall '08
- Machine Learning