<ul><li>Unsupervised multivariate time series anomaly detection (UMTSAD) is important in various domains.</li><li>Deep learning models based on the Transformer and self-attention mechanisms have shown impressive results in UMTSAD.</li><li>However, these models have limitations in generalizing to diverse anomaly situations without labeled data.</li><li>To address this, the proposed model, AMAD, integrates AutoMasked Attention for UMTSAD scenarios, providing a robust and adaptable solution.</li></ul>

AMAD: AutoMasked Attention for Unsupervised Multivariate Time Series Anomaly Detection

Discover more