Pruning Extensions and Efficiency Trade-Offs for Sustainable Time Series Classification

2026-04-09

AI Summary

This study addresses the lack of systematic understanding regarding the trade-offs among model performance, dataset characteristics, and hardware energy efficiency in time series classification (TSC). To bridge this gap, the authors establish a unified evaluation framework, introduce a theoretically bounded pruning strategy to optimize the prominent hybrid models Hydra and Quant, and design a novel prunable architecture named Hydrant that effectively balances predictive accuracy and resource efficiency. Experimental results across 20 MONSTER benchmark datasets demonstrate that the proposed approach can reduce energy consumption by up to 80% with an average accuracy degradation of less than 5%. This work provides the first systematic validation of efficient, sustainable, and reproducible practices for TSC.

Abstract

Time series classification (TSC) enables important use cases, however lacks a unified understanding of performance trade-offs across models, datasets, and hardware. While resource awareness has grown in the field, TSC methods have not yet been rigorously evaluated for energy efficiency. This paper introduces a holistic evaluation framework that explicitly explores the balance of predictive performance and resource consumption in TSC. To boost efficiency, we apply a theoretically bounded pruning strategy to leading hybrid classifiers - Hydra and Quant - and present Hydrant, a novel, prunable combination of both. With over 4000 experimental configurations across 20 MONSTER datasets, 13 methods, and three compute setups, we systematically analyze how model design, hyperparameters, and hardware choices affect practical TSC performance. Our results showcase that pruning can significantly reduce energy consumption by up to 80% while maintaining competitive predictive quality, usually costing the model less than 5% of accuracy. The proposed methodology, experimental results, and accompanying software advance TSC toward sustainable and reproducible practice.

Problem

• time series classification
• energy efficiency
• performance trade-offs
• resource consumption
• sustainable machine learning

Innovation

• pruning
• time series classification
• energy efficiency
• Hydrant
• sustainable machine learning