Robust PCA for Anomaly Detection and Data Imputation in Seasonal Time Series
This work addresses anomaly detection and data imputation in seasonal time series, which is incremental as it builds on existing RPCA frameworks with an online adaptation.
The authors tackled the problem of recovering low-rank and sparse matrices from seasonal time series data by proposing a robust PCA framework, including an online version for streaming data, and empirically demonstrated its effectiveness in anomaly detection and data imputation compared to other RPCA methods.
We propose a robust principal component analysis (RPCA) framework to recover low-rank and sparse matrices from temporal observations. We develop an online version of the batch temporal algorithm in order to process larger datasets or streaming data. We empirically compare the proposed approaches with different RPCA frameworks and show their effectiveness in practical situations.