NEFeb 22, 2018

Complex-valued Neural Networks with Non-parametric Activation Functions

arXiv:1802.08026v196 citations
Originality Incremental advance
AI Analysis

This work addresses a domain-specific problem for researchers and practitioners in fields like signal processing where complex-valued data is common, representing an incremental advance by adapting existing kernel methods to the complex domain.

The paper tackled the challenge of designing flexible activation functions for complex-valued neural networks (CVNNs) by proposing the first fully complex, non-parametric activation function based on kernel expansions, which was validated in experiments on prediction and channel equalization tasks.

Complex-valued neural networks (CVNNs) are a powerful modeling tool for domains where data can be naturally interpreted in terms of complex numbers. However, several analytical properties of the complex domain (e.g., holomorphicity) make the design of CVNNs a more challenging task than their real counterpart. In this paper, we consider the problem of flexible activation functions (AFs) in the complex domain, i.e., AFs endowed with sufficient degrees of freedom to adapt their shape given the training data. While this problem has received considerable attention in the real case, a very limited literature exists for CVNNs, where most activation functions are generally developed in a split fashion (i.e., by considering the real and imaginary parts of the activation separately) or with simple phase-amplitude techniques. Leveraging over the recently proposed kernel activation functions (KAFs), and related advances in the design of complex-valued kernels, we propose the first fully complex, non-parametric activation function for CVNNs, which is based on a kernel expansion with a fixed dictionary that can be implemented efficiently on vectorized hardware. Several experiments on common use cases, including prediction and channel equalization, validate our proposal when compared to real-valued neural networks and CVNNs with fixed activation functions.

Code Implementations2 repos
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes