A General Framework for Hypercomplex-valued Extreme Learning Machines

Abstract

This paper aims to establish a framework for extreme learning machines (ELMs) on general hypercomplex algebras. Hypercomplex neural networks are machine learning models that feature higher-dimension numbers as parameters, inputs, and outputs. Firstly, we review broad hypercomplex algebras and show a framework to operate in these algebras through real-valued linear algebra operations in a robust manner. We proceed to explore a handful of well-known four-dimensional examples. Then, we propose the hypercomplex-valued ELMs and derive their learning using a hypercomplex-valued least-squares problem. Finally, we compare real and hypercomplex-valued ELM models' performance in an experiment on time-series prediction and another on color image autoencoding. The computational experiments highlight the excellent performance of hypercomplex-valued ELMs to treat high-dimensional data, including models based on unusual hypercomplex algebras.

CDRec: Cayley-Dickson Recommender

“…In computer vision, quaternion convolutional neural networks (QCNNs) obtain better performance than traditional real-valued models on image classification [25,43]. Hypercomplex-valued extreme learning machines (HELMs) make great progress in color image auto-encoding [36,37]. Hypercomplex representations are also useful for enhancing the performance of the transformer model for natural language processing [31].…”