Artificial Intelligence for Big Data

Anomaly detection 

As we start gathering data from heterogeneous systems, there is a pattern that is established in terms of data volume, structure, information content, and velocity of the data. This pattern remains consistent during standard operating conditions and there can be expected surges or changes in the patterns. For example, an online retailer can expect more orders during the holiday season and this event does not count as an anomaly. When there is an unanticipated change in the regular pattern of data in terms of volume, velocity, and variety, the anomaly detection component triggers an alert and notification. One of the important characteristics of a greatly evolved and reliable anomaly detection component is that it is able ...