SPOCC: Scalable POssibilistic Classifier Combination -- toward robust aggregation of classifiers

Abstract

We investigate a problem in which each member of a group of learners is trained separately to solve the same classification task. Each learner has access to a training dataset (possibly with overlap across learners) but each trained classifier can be evaluated on a validation dataset. We propose a new approach to aggregate the learner predictions in the possibility theory framework. For each classifier prediction, we build a possibility distribution assessing how likely the classifier prediction is correct using frequentist probabilities estimated on the validation set. The possibility distributions are aggregated using an adaptive t-norm that can accommodate dependency and poor accuracy of the classifier predictions. We prove that the proposed approach possesses a number of desirable classifier combination robustness properties.

References in corpus (6)

• Proceedings of the Fifteenth Conference on Uncertainty in Artificial Intelligence (1999)
• Federated Learning with Unbiased Gradient Aggregation and Controllable Meta Updating
• COBRA: A Combined Regression Strategy
• Possibilistic logic bases and possibilistic graphs
• On Single Source Robustness in Deep Fusion Models
• Online Hierarchical Clustering Approximations

Cited by in corpus (1)

• Test-time Collective Prediction