Difference between revisions of "Thread:User talk:AW/KNN/Mahalanobis Distance/reply (6)"

From Robowiki
Jump to navigation Jump to search
 
 
Line 1: Line 1:
Mahalanobis distance can be combined with KNN search. Swap point distance calculation from euclidean or mahalanobis and you have it.
+
Mahalanobis distance can be combined with KNN search. Swap point distance calculation from euclidean to mahalanobis and you have it.
  
 
Another way of viewing mahalanobis distance is it rotates the data space, or have "diagonal" weights. Which is good when classifiers are not completely independent from each other.
 
Another way of viewing mahalanobis distance is it rotates the data space, or have "diagonal" weights. Which is good when classifiers are not completely independent from each other.
  
 
i.e. lateral velocity and vertical velocity. If lateral velocity is 8, vertical velocity is always 0. Coordinates where both velocities are 8 are always empty. Mahalanobis distance then uses a "diagonal" weight so 5.6 lateral velocity and 5.6 vertical velocity (5.6, 5.6) is normalized as (8, 8). While (8, 0) is still normalized as (8, 0).
 
i.e. lateral velocity and vertical velocity. If lateral velocity is 8, vertical velocity is always 0. Coordinates where both velocities are 8 are always empty. Mahalanobis distance then uses a "diagonal" weight so 5.6 lateral velocity and 5.6 vertical velocity (5.6, 5.6) is normalized as (8, 8). While (8, 0) is still normalized as (8, 0).

Latest revision as of 22:24, 19 October 2012

Mahalanobis distance can be combined with KNN search. Swap point distance calculation from euclidean to mahalanobis and you have it.

Another way of viewing mahalanobis distance is it rotates the data space, or have "diagonal" weights. Which is good when classifiers are not completely independent from each other.

i.e. lateral velocity and vertical velocity. If lateral velocity is 8, vertical velocity is always 0. Coordinates where both velocities are 8 are always empty. Mahalanobis distance then uses a "diagonal" weight so 5.6 lateral velocity and 5.6 vertical velocity (5.6, 5.6) is normalized as (8, 8). While (8, 0) is still normalized as (8, 0).