Novelty Detection in Time Series Data using Ideas from Immunology

Detecting anomalies in time series data is a problem of great practical interest in many manufacturing and signal processing applications. This paper presents a novelty detection algorithm inspired by the negative-selection mechanism of the immune system, which discriminates between self and other. Here self is deened to be normal data patterns and non-self is any deviation exceeding an allowable variation. Experiments with this novelty detection algorithm are reported for two data sets-simulated cutting dynamics in a milling operation and a synthetic signal. The results of the experiments exhibiting the performance of the algorithm in detecting novel patterns are reported.