The problem of learning patterns from sensor data arises in many applications, including e-commerce, smart environments, video surveillance, network analysis, human-robot interaction, ambient assisted living, and so on. We focus on detecting patterns that deviate from regular behaviors and might represent a security risk, health problem, or any other abnormal behavior contingency.

In other words, deviant behavior is a data pattern that either does not conform to the expected behavior (anomalous behavior) or matches a previously defined unwanted behavior (suspicious behavior). Deviant behavior patterns are also referred to as outliers, exceptions, peculiarities, surprises, misuse, and so on. Such patterns occur relatively infrequently; however, when they do occur, their consequences can be quite dramatic, and often negatively so. Typical examples include credit card fraud, cyber intrusions, and industrial damage. In e-commerce, fraud is estimated to cost merchants more than $200 billion a year; in healthcare, fraud is estimated to cost taxpayers $60 billion a year; for banks, the cost is over $12 billion.