Divide and conquer: neuroevolution for multiclass classification

Title

Divide and conquer: neuroevolution for multiclass classification

Creator

Schrum, Jacob

McDonnell, Tyler

Andoni, Sari

Bonab, Elmira

Cheng, Sheila

Goode, Jimmie

Moore, Keith

Sellers, Gavin

Choi, Jun-Hwan

Date

2018-12-10

Date Available

2018-12-10

Date Issued

2018-07

Identifier

Tyler McDonnell, Sari Andoni, Elmira Bonab, Sheila Cheng, Jun-Hwan Choi, Jimmie Goode, Keith Moore, Gavin Sellers, and Jacob Schrum. 2018. Divide and conquer: neuroevolution for multiclass classification. In Proceedings of the Genetic and Evolutionary Computation Conference (GECCO '18), Hernan Aguirre (Ed.). ACM, New York, NY, USA, 474-481. DOI: https://doi.org/10.1145/3205455.3205476

uri

https://collections.southwestern.edu/s/suscholar/item/247

Abstract

Neuroevolution is a powerful and general technique for evolving the structure and weights of artificial neural networks. Though neuroevolutionary approaches such as NeuroEvolution of Augmenting Topologies (NEAT) have been successfully applied to various problems including classification, regression, and reinforcement learning problems, little work has explored application of these techniques to larger-scale multiclass classification problems. In this paper, NEAT is evaluated in several multiclass classification problems, and then extended via two ensemble approaches: One-vs-All and One-vs-One. These approaches decompose multiclass classification problems into a set of binary classification problems, in which each binary problem is solved by an instance of NEAT. These ensemble models exhibit reduced variance and increasingly superior accuracy as the number of classes increases. Additionally, higher accuracy is achieved early in training, even when artificially constrained for the sake of fair comparison with standard NEAT. However, because the approach can be trivially distributed, it can be applied quickly at large scale to solve real problems. In fact, these approaches are incorporated into DarwinTM, an enterprise automatic machine learning solution that also incorporates various other algorithmic enhancements to NEAT. The resulting complete system has proven robust to a wide variety of client datasets.

Language

English

Publisher

GECCO '18 Proceedings of the Genetic and Evolutionary Computation Conference

Subject

Computing methodologies

Machine learning

Machine learning approaches

Neural networks

Bio-inspired approaches

Generative and developmental approaches

Type

Article