Journal
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
Volume 32, Issue 9, Pages 4253-4266Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TNNLS.2020.3017675
Keywords
Bandwidth; Anomaly detection; Kernel; Estimation; Partitioning algorithms; Data models; Real-time systems; Anomaly detection; bandwidth selection; kernel density estimation; online; regret analysis
Categories
Funding
- Scientific and Technological Research Council of Turkey [118E268, 117E153]
Ask authors/readers for more resources
The proposed algorithm is an unsupervised anomaly detection algorithm that can handle sequential data from complex distributions online with strong performance guarantees. By constructing a partitioning tree and training online kernel density estimators, the algorithm sequentially produces a final density estimation for anomaly detection by comparing it with a data-adaptive threshold. The computational complexity is linear in both tree depth and data length, and significant improvements in accuracy compared to state-of-the-art techniques have been observed in experiments.
We propose a novel unsupervised anomaly detection algorithm that can work for sequential data from any complex distribution in a truly online framework with mathematically proven strong performance guarantees. First, a partitioning tree is constructed to generate a doubly exponentially large hierarchical class of observation space partitions, and every partition region trains an online kernel density estimator (KDE) with its own unique dynamical bandwidth. At each time, the proposed algorithm optimally combines the class estimators to sequentially produce the final density estimation. We mathematically prove that the proposed algorithm learns the optimal partition with kernel bandwidths that are optimized in both region-specific and time-varying manner. The estimated density is then compared with a data-adaptive threshold to detect anomalies. Overall, the computational complexity is only linear in both the tree depth and data length. In our experiments, we observe significant improvements in anomaly detection accuracy compared with the state-of-the-art techniques.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available