ReLATE: Resilient Learner Selection for Multivariate Time-Series Classification Against Adversarial Attacks
Cagla Ipek Kocal, Onat Gungor, Aaron Tartz, Tajana Rosing, Baris Aksanli
TL;DR
The paper addresses robust multivariate time-series classification in streaming contexts where adversarial perturbations are possible, while minimizing retraining costs. It introduces ReLATE, a four-module framework that builds a Performance Benchmark Database, applies adversarial tests, learns dataset embeddings via a lightweight CNN, and selects the top three models from the most similar dataset for a new arrival. Empirical results show substantial overhead reductions (about 81.2% on average) with performance close to an Oracle (within about 4%), and improved resilience over random baselines under multiple attack scenarios. This similarity-based resilient learner selection enables fast, scalable, and robust MTSC in real-time settings facing adversarial threats.
Abstract
Minimizing computational overhead in time-series classification, particularly in deep learning models, presents a significant challenge. This challenge is further compounded by adversarial attacks, emphasizing the need for resilient methods that ensure robust performance and efficient model selection. We introduce ReLATE, a framework that identifies robust learners based on dataset similarity, reduces computational overhead, and enhances resilience. ReLATE maintains multiple deep learning models in well-known adversarial attack scenarios, capturing model performance. ReLATE identifies the most analogous dataset to a given target using a similarity metric, then applies the optimal model from the most similar dataset. ReLATE reduces computational overhead by an average of 81.2%, enhancing adversarial resilience and streamlining robust model selection, all without sacrificing performance, within 4.2% of Oracle.