Incorporating Fuzzy ARTMAP Network with Spiking Neural Networks

Abstract

In this paper, we have incorporated the Fuzzy ARTMAP classification module into the Spiking Neural Networks (SNNs). This new network can have multiple hidden layers, a supervised corrected layer, and an ARTMAP module. The objective of the supervised layer is to correct the spikes given by the Spike-timing-dependent plasticity (STDP) layer in a supervised manner. Then the resulting corrected spikes are fed into an ARTMAP network. When an output spiking train code is similar to one of the stored ART codes. It modifies that code to incorporate the new input into that code category. This combination required a new training algorithm. The hidden layers are trained using the popular STDP. This unsupervised learning is followed by a supervised layer where new supervised weight update formulas are presented. This new combination is compared to other spiking algorithms using datasets from the UCI repository. It is shown that SNN_ARTMAP provides better recognition results than other similar spiking supervised learning algorithms.

Received: 14 April 2025 | Revised: 4 June 2025 | Accepted: 22 August 2025

Conflicts of Interest

The author declares that he has no conflicts of interest to this work.

Data Availability Statement

The data that support the findings of this study are openly available in the UC Irvine Machine Learning Repository at https://archive.ics.uci.edu/.

Author Contribution Statement

Issam Dagher: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration.