SpikeLite: Lightweight Spiking Neural Networks for Time-Series Forecasting

2026-09-28

AI Summary

This study addresses the structural bloat in existing spiking neural network (SNN) predictors, which arises from an excessive pursuit of accuracy and undermines their inherent advantages in lightweight design and energy efficiency. To this end, we propose SpikeLite, a framework incorporating a frequency-selective encoder and a sparse channel attention module. Methodologically, the approach exploits the low-pass filtering properties of leaky integrate-and-fire (LIF) neurons to reorganize frequency-domain components, while employing binary masks to suppress redundant interactions and support channel-independent pathways, thereby reducing computational overhead. Experimental results demonstrate that the proposed framework achieves state-of-the-art comprehensive performance across multiple benchmarks while significantly lowering energy consumption, validating the potential of SNNs for efficient time series forecasting.

Abstract

Spiking neural networks (SNNs) offer an energy-efficient paradigm for time-series forecasting through spike-driven computation. However, recent SNN forecasters often pursue higher accuracy through increasingly complex attention mechanisms, or specialized neuronal dynamics, weakening the lightweight motivation of SNNs. We introduce SpikeLite, a spiking forecasting framework built around two modules: a Frequency-Selective Spiking Encoder (FSSE) for frequency-sensitive temporal encoding and a Sparse Spiking Channel Attention (SSCA) module for selective cross-channel interaction. FSSE exploits the low-pass filtering behavior of LIF dynamics to reorganize each input sequence into frequency-sensitive components while collectively preserving the input at the decomposition stage. SSCA then learns a binary mask from encoded channel representations and uses it to selectively exchange information within spike-driven self-attention, retaining informative cross-channel interactions while suppressing redundant ones. When explicit channel interaction is unnecessary, SpikeLite uses the lighter FSSE-only channel-independent path. Experiments under the SeqSNN and SpikF protocols cover four standard multivariate and eight long-term forecasting benchmarks. SpikeLite achieves the best aggregate performance under both protocols, with an average $R^2$ of 0.790 and RSE of 0.440, and lowest average MSE/MAE of 0.343/0.345 in long-term forecasting. Moreover, evaluation on the ECL dataset shows that SpikeLite achieves the lowest reported energy consumption, further demonstrating its potential for energy-efficient time-series forecasting.

Problem

• Spiking Neural Networks
• Time-Series Forecasting
• Lightweight
• Energy Efficiency

Innovation

• Spiking Neural Networks
• Time-Series Forecasting
• Frequency-Selective Spiking Encoder
• Sparse Spiking Channel Attention
• Energy-Efficient