Supervised learning for change-point detection. (English) Zbl 1128.68424

Summary: The detection of changes in the distribution of process variables is referred to as the change-point problem. Existing methods focus on detecting a single (or few) change point in a univariate (or low-dimensional) process. We consider the important high-dimensional multivariate case with multiple change points and without an assumed distribution. In this work the problem is transformed into a supervised learning problem with time as the output response and the process variables as inputs. Our focus is to identify the subset of variables that change. This important, practical scenario is analysed through a supervised learner with a variable importance measure that is used to identify the variables that change among hundreds of variables. Simulated cases are discussed in the paper to verify the proposed method. Moreover, the same data sets are compared with a multivariate exponentially weighted moving average control chart and the advantages of the supervised learner are illustrated.


68T05 Learning and adaptive systems in artificial intelligence
62P30 Applications of statistics in engineering and industry; control charts
Full Text: DOI


[1] Belisle P, Biometrics 54 pp 1998– (1998)
[2] Bhattacharya PK, J. multivar. Anal. 23 pp 1987– (1987)
[3] Breiman L, Mach. Learn. 45 pp 2001– (2001)
[4] Breiman L, Classification and Regression Trees (1984)
[5] Carlstein E, Ann. Statist. 16 pp 1988– (1988)
[6] Csörgö M, J. statist. Plann. Infer. 17 pp 1987– (1987)
[7] Davison AD, Bootstrap Methods and Their Application (1997)
[8] Fu Y-X, Biometrika 77 pp 1990– (1990)
[9] Hinkley DV, Biometrika 57 pp 1970– (1970)
[10] Hinkley DV, Biometrika 58 pp 509– (1971) · Zbl 0254.62019
[11] Hotelling HH, Techniques of Statistical Analysis (1947)
[12] Hwang W, Inst. ind. Engrs Trans. (2004)
[13] Lavielle M, IEEE Trans. Signal Process. 41 pp 1993– (1993) · Zbl 0775.68016
[14] Lee CB, Statist. Probab. Lett. 27 pp 1996– (1996)
[15] Lowry CA, Technometrics 34 pp 1992– (1992)
[16] Pievatolo A, Appl. Statist. 49 pp 2000– (2000)
[17] Runger GC, Qual. Reliabil. Int. 20 pp 2004– (2004)
[18] Taylor AR, Int. J. Adapt. Control Signal Process 17 pp 2003– (2003)
[19] Tuv E, Feature selection using ensemble-based ranking against artificial contrasts (2005)
[20] Xiong Y, Patt. Recogn. 37 pp 2004– (2004)
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.