FEATHer: Fourier-Efficient Adaptive Temporal Hierarchy Forecaster for Time-Series Forecasting

2026-01-16

AI Summary

This work addresses the challenge of achieving high-accuracy, low-latency, and memory-efficient long-term time series forecasting on industrial edge devices—such as PLCs and microcontrollers—under extreme parameter constraints. The authors propose an ultra-lightweight model featuring a novel frequency-domain adaptive hierarchical architecture that integrates multi-scale Fourier decomposition, projection-depthwise convolution-projection modules, spectrum-aware gating, and periodic sparse reconstruction kernels, deliberately avoiding attention mechanisms and recurrent structures. With only 400 parameters, the model achieves an average rank of 2.05 across eight benchmark datasets and secures first place in 60 evaluation settings, significantly outperforming existing lightweight approaches and demonstrating the feasibility of highly efficient long-term forecasting at the edge.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Natural Language Processing: Learning & Optimization for NLP

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Time-series forecasting is fundamental in industrial domains like manufacturing and smart factories. As systems evolve toward automation, models must operate on edge devices (e.g., PLCs, microcontrollers) with strict constraints on latency and memory, limiting parameters to a few thousand. Conventional deep architectures are often impractical here. We propose the Fourier-Efficient Adaptive Temporal Hierarchy Forecaster (FEATHer) for accurate long-term forecasting under severe limits. FEATHer introduces: (i) ultra-lightweight multiscale decomposition into frequency pathways; (ii) a shared Dense Temporal Kernel using projection-depthwise convolution-projection without recurrence or attention; (iii) frequency-aware branch gating that adaptively fuses representations based on spectral characteristics; and (iv) a Sparse Period Kernel reconstructing outputs via period-wise downsampling to capture seasonality. FEATHer maintains a compact architecture (as few as 400 parameters) while outperforming baselines. Across eight benchmarks, it achieves the best ranking, recording 60 first-place results with an average rank of 2.05. These results demonstrate that reliable long-range forecasting is achievable on constrained edge hardware, offering a practical direction for industrial real-time inference.

Problem

• time-series forecasting
• edge devices
• resource-constrained
• long-term forecasting
• industrial automation

Innovation

• Fourier decomposition
• ultra-lightweight forecasting
• frequency-aware gating
• sparse period kernel
• edge time-series forecasting