Add to Chrome
Add to Firefox
Add to EdgeTransNAS-TSAD: Harnessing Transformers for Multi-Objective Neural Architecture Search in Time Series Anomaly Detection
Dec 03, 2023
The surge in real-time data collection across various industries has underscored the need for advanced anomaly detection in both univariate and multivariate time series data. Traditional methods, while comprehensive, often struggle to capture the complex interdependencies in such data. This paper introduces TransNAS-TSAD, a novel framework that synergizes transformer architecture with neural architecture search (NAS), enhanced through NSGA-II algorithm optimization. This innovative approach effectively tackles the complexities of both univariate and multivariate time series, balancing computational efficiency with detection accuracy. Our evaluation reveals that TransNAS-TSAD surpasses conventional anomaly detection models, demonstrating marked improvements in diverse data scenarios. We also propose the Efficiency-Accuracy-Complexity Score (EACS) as a new metric for assessing model performance, emphasizing the crucial balance between accuracy and computational resources. TransNAS-TSAD sets a new benchmark in time series anomaly detection, offering a versatile, efficient solution for complex real-world applications. This research paves the way for future developments in the field, highlighting its potential in a wide range of industry applications.
