Optimal Binary Classifier Aggregation for General Losses
–Neural Information Processing Systems
We address the problem of aggregating an ensemble of predictors with known loss bounds in a semi-supervised binary classification setting, to minimize prediction loss incurred on the unlabeled data. We find the minimax optimal predictions for a very general class of loss functions including all convex and many non-convex losses, extending a recent analysis of the problem for misclassification error. The result is a family of semi-supervised ensemble aggregation algorithms which are as efficient as linear learning by convex optimization, but are minimax optimal without any relaxations. Their decision rules take a form familiar in decision theory - applying sigmoid functions to a notion of ensemble margin - without the assumptions typically made in margin-based learning.
Neural Information Processing Systems
Mar-12-2024, 18:59:38 GMT