METHODS FOR DETECTING AND INTERPRETING DATA ANOMALIES, AND RELATED SYSTEMS AND DEVICES

摘要

An anomaly detection method comprising: obtaining a data set comprising plurality of data samples, each of the plurality of data samples associated with respective values for a set of features;identifying a respective data type of each of the features; automatically generating an anomaly detection blueprint based on the respective data types of one or more of the features, the anomaly detection blueprint comprising a machine-executable module encoding an anomaly detection procedure, the anomaly detection procedure including a plurality of tasks, wherein one or more of the tasks are selected based on a size of the data set and/or the respective data types of the one or more features; and executing the machine-executable module, thereby performing the anomaly detection procedure, wherein performing the anomaly detection procedure includes identifying a subset of the plurality of data samples as a set of anomalous data samples.