Random Classification Noise Defeats All Convex Potential Boosters thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Random Classification Noise Defeats All Convex Potential Boosters

Published on Aug 06, 20083564 Views

A broad class of boosting algorithms can be interpreted as performing coordinate-wise gradient descent to minimize some potential function of the margins of a data set. This class includes AdaBoost, L