Multiple changepoint detection in categorical data streams

Abstract

The need for efficient tools is pressing in the era of big data, particularly in streaming data applications. As data streams are ubiquitous, the ability to accurately detect multiple changepoints, without affecting the continuous flow of data, is an important issue. Change detection for categorical data streams is understudied, and existing work commonly introduces fixed control parameters while providing little insight into how they may be chosen. This is ill-suited to the streaming paradigm, motivating the need for an approach that introduces few parameters which may be set without requiring any prior knowledge of the stream. This paper introduces such a method, which can accurately detect changepoints in categorical data streams with fixed storage and computational requirements. The detector relies on the ability to adaptively monitor the category probabilities of a multinomial distribution, where temporal adaptivity is introduced using forgetting factors. A novel adaptive threshold is also developed which can be computed given a desired false positive rate. This method is then compared to sequential and nonsequential change detectors in a large simulation study which verifies the usefulness of our approach. A real data set consisting of nearly 40 million events from a computer network is also investigated.

Identifying irregular activity sequences: an application to passive household monitoring

“…Höhle (2010) presents methods to detect changes in categorical time series using multi-categorical regression models where they monitor the category probabilities seen within the data. Recently, Plasse and Adams (2019) propose to detect changepoints in multiple categorical data streams that monitor the category probabilities of a multinomial distribution. Generally these approaches focus on estimating changes in the probabilities of sensors activating within a single day but will not detect changes in the order of the events.…”

Lightweight Automated Feature Monitoring for Data Streams

“…In Ref. [26] the authors approximate the multinomial distribution of a stream of values of a categorical feature by using a relative frequencies histogram. Their change detection is based on the Kullback Leibler (KL) divergence between static and adaptive estimates of the multinomial densities.…”

Impact of Variable Size Chunk Data on Classifier Performance using Ensemble Techniques

“…Muhammad et al [15] proposed a Monte Carlo tree search (MTCS) method for retrieving features from very high dimensional data. Multiple change detection methods have been proposed for categorical data stream by Joshua Plasse and Niali Adams [16]. Kyosuke Nishida and Koichiro Yamauchi [17] developed a STEPD algorithm which uses a statistical test for equal proportion for online environment.…”