HIT-ROCKET: Hadamard-vector Inner-product Transformer for ROCKET

2025-11-03

AI Summary

To address the high computational complexity of state-of-the-art time series classification methods and the limited expressiveness–efficiency trade-off of lightweight models (e.g., ROCKET), this paper proposes a Hadamard-based vectorized convolutional kernel transformation for feature extraction. Leveraging the orthogonality and fast computability of Hadamard matrices, we design scalable, multi-scale random convolutional kernels that remain fully compatible with the ROCKET framework. Our method integrates Hadamard inner-product transformation, random projection, and temporal mapping to significantly enhance feature discriminability while reducing computational overhead—enabling deployment on ultra-low-power devices. Evaluated on the UCR benchmark, it achieves ≥5% improvement in F1-score over ROCKET and reduces training time by 50% compared to miniROCKET, establishing new state-of-the-art performance.

Technology Category

• Machine Learning: Kernel Methods
• Computer Vision: Adversarial Attacks & Robustness
• Intelligent Robots: State Estimation

Application Category

• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs

Abstract

Time series classification holds broad application value in communications, information countermeasures, finance, and medicine. However, state-of-the-art (SOTA) methods-including HIVE-COTE, Proximity Forest, and TS-CHIEF-exhibit high computational complexity, coupled with lengthy parameter tuning and training cycles. In contrast, lightweight solutions like ROCKET (Random Convolutional Kernel Transform) offer greater efficiency but leave substantial room for improvement in kernel selection and computational overhead. To address these challenges, we propose a feature extraction approach based on Hadamard convolutional transform, utilizing column or row vectors of Hadamard matrices as convolution kernels with extended lengths of varying sizes. This enhancement maintains full compatibility with existing methods (e.g., ROCKET) while leveraging kernel orthogonality to boost computational efficiency, robustness, and adaptability. Comprehensive experiments on multi-domain datasets-focusing on the UCR time series dataset-demonstrate SOTA performance: F1-score improved by at least 5% vs. ROCKET, with 50% shorter training time than miniROCKET (fastest ROCKET variant) under identical hyperparameters, enabling deployment on ultra-low-power embedded devices. All code is available on GitHub.

Problem

• Reducing computational complexity in time series classification methods
• Improving kernel selection efficiency for lightweight ROCKET algorithms
• Enhancing performance while maintaining compatibility with existing approaches

Innovation

• Uses Hadamard matrix vectors as convolution kernels
• Enhances computational efficiency with kernel orthogonality
• Maintains full compatibility with existing ROCKET methods