TY - GEN

T1 - Locating changes in highly-dependent data with an unknown number of change-points

AU - Khaleghi, Azedeh

AU - Ryabko, Daniil

PY - 2012

Y1 - 2012

N2 - The problem of multiple change point estimation is considered for sequences withunknown number of change points. A consistency framework is suggested thatis suitable for highly dependent time-series, and an asymptotically consistent algorithm is proposed. In order for the consistency to be established the only assumption required is that the data is generated by stationary ergodic time-seriesdistributions. No modeling, independence or parametric assumptions are made;the data are allowed to be dependent and the dependence can be of arbitrary form.The theoretical results are complemented with experimental evaluations.

AB - The problem of multiple change point estimation is considered for sequences withunknown number of change points. A consistency framework is suggested thatis suitable for highly dependent time-series, and an asymptotically consistent algorithm is proposed. In order for the consistency to be established the only assumption required is that the data is generated by stationary ergodic time-seriesdistributions. No modeling, independence or parametric assumptions are made;the data are allowed to be dependent and the dependence can be of arbitrary form.The theoretical results are complemented with experimental evaluations.

KW - Change Point Analysis

KW - Stationary Ergodic Processes

KW - Unsupervised Learning

KW - Consistency

M3 - Conference contribution/Paper

SN - 9781627480031

SP - 1

EP - 9

BT - Advances in Neural Information Processing Systems 25 (NIPS 2012)

A2 - Pereira, F.

A2 - Burges, C. J. C.

A2 - Bottou, L.

A2 - Weinberger, K. Q.

T2 - Neural Information Processing Systems (NIPS)

Y2 - 3 September 2012

ER -