Unsupervised Contextual Anomaly Detection using Joint Deep Variational Generative Models

Shulman, Yaniv

arXiv.org Machine Learning 

Often these processes result in highly dimensional data sets, with complex relationships within the data and exhibit stochastic behavior. Furthermore the anomalies by definition contain high self-information measure and therefore carry useful information about the underlying data generation process. There exist a number of similar definitions of what an anomaly is however in this paper the following definition is adopted [11]: 1. Anomalies are different from the norm in respect to their attributes.

Duplicate Docs Excel Report

None found

Similar Docs  Excel Report  more

None found