Consider a scenario with Dataset D as {T1, T2, ..., Tn} and Ti is a multivariate time-series of length mi as {X1, X2, ..., Xmi}. Here each record of the time-series Xi is a vector of attribute values {Xi1, Xi2, ..., Xia} where Xij is the value of jth attribute of ith record of a time series. Given this scenario is there a method to detect the anomalous time-series(or a set of anomalous time series, i.e the T's which are anomalous) from the set D?
Please look at the following below links which may help you in your analysis:
https://www.researchgate.net/publication/328183877_Multivariate-Time-Series-Driven_Real-time_Anomaly_Detection_Based_on_Bayesian_Network
https://www.osti.gov/servlets/purl/1514382
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2656052/
Thanks
There are four broad ways to detect anomaly in time series.
1.state space models- holtwinters, arima
2.Decomposition-STL method
3.Deep learning- recurrent neural network
4.Dimensionality reduction-RPCA, SOM etc.
- Badges
- Science topic
- Similar topics
- Statistics
- Time Series