Skip to yearly menu bar Skip to main content


Optimal Minimal Margin Maximization with Boosting

Alexander Mathiasen · Kasper Green Larsen · Allan Grønlund

Pacific Ballroom #263

Keywords: [ Ensemble Methods ]


Boosting algorithms iteratively produce linear combinations of more and more base hypotheses and it has been observed experimentally that the generalization error keeps improving even after achieving zero training error. One popular explanation attributes this to improvements in margins. A common goal in a long line of research, is to obtain large margins using as few base hypotheses as possible, culminating with the AdaBoostV algorithm by Rätsch and Warmuth [JMLR’05]. The AdaBoostV algorithm was later conjectured to yield an optimal trade-off between number of hypotheses trained and the minimal margin over all training points (Nie, Warmuth, Vishwanathan and Zhang [JMLR’13]). Our main contribution is a new algorithm refuting this conjecture. Furthermore, we prove a lower bound which implies that our new algorithm is optimal.

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