Tree Methods for Hierarchical Classification in Parallel

TL;DR

This paper introduces efficient parallel methods for hierarchical classification that transform batch scores and labels into ancestral paths using tensor operations, enabling scalable classification over large semantic trees.

Contribution

It presents novel tensor-based algorithms for hierarchical classification that operate efficiently on hardware accelerators, handling large trees with minimal computational overhead.

Findings

Transform batches of scores to ancestral paths with negligible computation.

Implemented methods on hardware accelerators for WordNet's large hierarchy.

Achieved minimal memory footprint of 0.04GB during processing.

Abstract

We propose methods that enable efficient hierarchical classification in parallel. Our methods transform a batch of classification scores and labels, corresponding to given nodes in a semantic tree, to scores and labels corresponding to all nodes in the ancestral paths going down the tree to every given node, relying only on tensor operations that execute efficiently on hardware accelerators. We implement our methods and test them on current hardware accelerators with a tree incorporating all English-language synsets in WordNet 3.0, spanning 117,659 classes in 20 levels of depth. We transform batches of scores and labels to their respective ancestral paths, incurring negligible computation and consuming only a fixed 0.04GB of memory over the footprint of data.