MTSA-SNN: A Multi-modal Time Series Analysis Model Based on Spiking Neural Network

TL;DR

This paper introduces MTSA-SNN, a novel spiking neural network model that effectively processes complex, multi-modal time series data by encoding, fusing, and analyzing temporal information with wavelet transforms, outperforming traditional methods.

Contribution

The paper presents a new multi-modal time series analysis model based on spiking neural networks, incorporating pulse encoding, joint learning, and wavelet transforms for improved temporal data analysis.

Findings

Achieved superior performance on three complex time-series tasks.

Effectively fuses multi-modal pulse signals for better information integration.

Enhances temporal analysis with wavelet transform operations.

Abstract

Time series analysis and modelling constitute a crucial research area. Traditional artificial neural networks struggle with complex, non-stationary time series data due to high computational complexity, limited ability to capture temporal information, and difficulty in handling event-driven data. To address these challenges, we propose a Multi-modal Time Series Analysis Model Based on Spiking Neural Network (MTSA-SNN). The Pulse Encoder unifies the encoding of temporal images and sequential information in a common pulse-based representation. The Joint Learning Module employs a joint learning function and weight allocation mechanism to fuse information from multi-modal pulse signals complementary. Additionally, we incorporate wavelet transform operations to enhance the model's ability to analyze and evaluate temporal information. Experimental results demonstrate that our method achieved superior performance on three complex time-series tasks. This work provides an effective event-driven approach to overcome the challenges associated with analyzing intricate temporal information. Access to the source code is available at https://github.com/Chenngzz/MTSA-SNN}{https://github.com/Chenngzz/MTSA-SNN