A Comparative Study of Outlier Detection in Large-Scale Data Using Data Mining Algorithms

Anomaly or outlier detection means detecting anomaly events from images, videos, online activities, data sets or state of the art. It is an identification of abnormal events using Sparse Coding (SC) Framework. Outlier detection is a primary step in data mining applications. The applications are intrusion detection, fraud detection, saliency detection and fault detection. Anomalous events occur infrequently, that may be detected from the consequences occurring in the images. Various techniques are used for outlier detection. The techniques are density-based outlier, correlation-based outlier, cluster-based outlier, neural networks, SC.