CALAD uses reconstruction-error-based channel relevance to construct contrastive samples in a transformer autoencoder framework, outperforming baselines on real-world multivariate time series anomaly detection tasks especially under distribution shift.
In: Proceed- ings of the 25th ACM SIGKDD international conference on knowledge discovery & data mining
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.LG 2years
2026 2representative citing papers
citing papers explorer
-
CALAD: Channel-Aware contrastive Learning for multivariate time series Anomaly Detection
CALAD uses reconstruction-error-based channel relevance to construct contrastive samples in a transformer autoencoder framework, outperforming baselines on real-world multivariate time series anomaly detection tasks especially under distribution shift.
- Conditional Attribution for Root Cause Analysis in Time-Series Anomaly Detection