SMORE: Similarity-based Hyperdimensional Domain Adaptation for Multi-Sensor Time Series Classification
This addresses performance degradation and computational inefficiency in IoT edge devices due to domain shifts, offering a resource-efficient solution, though it appears incremental as it builds on existing domain adaptation and hyperdimensional computing techniques.
The paper tackles the problem of distribution shift and resource constraints in multi-sensor time series classification for IoT applications by proposing SMORE, a domain adaptation algorithm that uses hyperdimensional computing, achieving on average 1.98% higher accuracy than SOTA DNN-based methods with 18.81x faster training and 4.63x faster inference.
Many real-world applications of the Internet of Things (IoT) employ machine learning (ML) algorithms to analyze time series information collected by interconnected sensors. However, distribution shift, a fundamental challenge in data-driven ML, arises when a model is deployed on a data distribution different from the training data and can substantially degrade model performance. Additionally, increasingly sophisticated deep neural networks (DNNs) are required to capture intricate spatial and temporal dependencies in multi-sensor time series data, often exceeding the capabilities of today's edge devices. In this paper, we propose SMORE, a novel resource-efficient domain adaptation (DA) algorithm for multi-sensor time series classification, leveraging the efficient and parallel operations of hyperdimensional computing. SMORE dynamically customizes test-time models with explicit consideration of the domain context of each sample to mitigate the negative impacts of domain shifts. Our evaluation on a variety of multi-sensor time series classification tasks shows that SMORE achieves on average 1.98% higher accuracy than state-of-the-art (SOTA) DNN-based DA algorithms with 18.81x faster training and 4.63x faster inference.