Online probabilistic label trees

TL;DR

This paper presents online probabilistic label trees (OPLTs), an efficient algorithm for online multi-label and multi-class classification that adapts to data without prior knowledge, suitable for challenging scenarios like one- or few-shot learning.

Contribution

Introduction of OPLTs, a novel online algorithm for label tree classification with low complexity and strong theoretical guarantees, capable of handling complex learning scenarios.

Findings

Effective in online multi-label and multi-class tasks

Performs well in one- and few-shot learning scenarios

Demonstrated strong empirical results across multiple datasets

Abstract

We introduce online probabilistic label trees (OPLTs), an algorithm that trains a label tree classifier in a fully online manner without any prior knowledge about the number of training instances, their features and labels. OPLTs are characterized by low time and space complexity as well as strong theoretical guarantees. They can be used for online multi-label and multi-class classification, including the very challenging scenarios of one- or few-shot learning. We demonstrate the attractiveness of OPLTs in a wide empirical study on several instances of the tasks mentioned above.