NEWMA: A New Method for Scalable Model-Free Online Change-Point Detection. Keriven, N., Garreau, D., & Poli, I. IEEE Transactions on Signal Processing, 68:3515–3528, 2020. Conference Name: IEEE Transactions on Signal Processingdoi abstract bibtex We consider the problem of detecting abrupt changes in the distribution of a multi-dimensional time series, with limited computing power and memory. In this paper, we propose a new, simple method for model-free online change-point detection that relies only on fast and light recursive statistics, inspired by the classical Exponential Weighted Moving Average algorithm (EWMA). The proposed idea is to compute two EWMA statistics on the stream of data with different forgetting factors, and to compare them. By doing so, we show that we implicitly compare recent samples with older ones, without the need to explicitly store them. Additionally, we leverage Random Features (RFs) to efficiently use the Maximum Mean Discrepancy as a distance between distributions, furthermore exploiting recent optical hardware to compute high-dimensional RFs in near constant time. We show that our method is significantly faster than usual non-parametric methods for a given accuracy.
@article{keriven_newma_2020,
title = {{NEWMA}: {A} {New} {Method} for {Scalable} {Model}-{Free} {Online} {Change}-{Point} {Detection}},
volume = {68},
issn = {1941-0476},
shorttitle = {{NEWMA}},
doi = {10.1109/TSP.2020.2990597},
abstract = {We consider the problem of detecting abrupt changes in the distribution of a multi-dimensional time series, with limited computing power and memory. In this paper, we propose a new, simple method for model-free online change-point detection that relies only on fast and light recursive statistics, inspired by the classical Exponential Weighted Moving Average algorithm (EWMA). The proposed idea is to compute two EWMA statistics on the stream of data with different forgetting factors, and to compare them. By doing so, we show that we implicitly compare recent samples with older ones, without the need to explicitly store them. Additionally, we leverage Random Features (RFs) to efficiently use the Maximum Mean Discrepancy as a distance between distributions, furthermore exploiting recent optical hardware to compute high-dimensional RFs in near constant time. We show that our method is significantly faster than usual non-parametric methods for a given accuracy.},
journal = {IEEE Transactions on Signal Processing},
author = {Keriven, Nicolas and Garreau, Damien and Poli, Iacopo},
year = {2020},
note = {Conference Name: IEEE Transactions on Signal Processing},
keywords = {Brain modeling, Change detection algorithms, Computational modeling, Hilbert space, Kernel, Microsoft Windows, Radio frequency, Signal processing algorithms, Streaming media, method of moments, optical computing},
pages = {3515--3528},
}
Downloads: 0
{"_id":"BQt4eL5HLuc4R27Pg","bibbaseid":"keriven-garreau-poli-newmaanewmethodforscalablemodelfreeonlinechangepointdetection-2020","author_short":["Keriven, N.","Garreau, D.","Poli, I."],"bibdata":{"bibtype":"article","type":"article","title":"NEWMA: A New Method for Scalable Model-Free Online Change-Point Detection","volume":"68","issn":"1941-0476","shorttitle":"NEWMA","doi":"10.1109/TSP.2020.2990597","abstract":"We consider the problem of detecting abrupt changes in the distribution of a multi-dimensional time series, with limited computing power and memory. In this paper, we propose a new, simple method for model-free online change-point detection that relies only on fast and light recursive statistics, inspired by the classical Exponential Weighted Moving Average algorithm (EWMA). The proposed idea is to compute two EWMA statistics on the stream of data with different forgetting factors, and to compare them. By doing so, we show that we implicitly compare recent samples with older ones, without the need to explicitly store them. Additionally, we leverage Random Features (RFs) to efficiently use the Maximum Mean Discrepancy as a distance between distributions, furthermore exploiting recent optical hardware to compute high-dimensional RFs in near constant time. We show that our method is significantly faster than usual non-parametric methods for a given accuracy.","journal":"IEEE Transactions on Signal Processing","author":[{"propositions":[],"lastnames":["Keriven"],"firstnames":["Nicolas"],"suffixes":[]},{"propositions":[],"lastnames":["Garreau"],"firstnames":["Damien"],"suffixes":[]},{"propositions":[],"lastnames":["Poli"],"firstnames":["Iacopo"],"suffixes":[]}],"year":"2020","note":"Conference Name: IEEE Transactions on Signal Processing","keywords":"Brain modeling, Change detection algorithms, Computational modeling, Hilbert space, Kernel, Microsoft Windows, Radio frequency, Signal processing algorithms, Streaming media, method of moments, optical computing","pages":"3515–3528","bibtex":"@article{keriven_newma_2020,\n\ttitle = {{NEWMA}: {A} {New} {Method} for {Scalable} {Model}-{Free} {Online} {Change}-{Point} {Detection}},\n\tvolume = {68},\n\tissn = {1941-0476},\n\tshorttitle = {{NEWMA}},\n\tdoi = {10.1109/TSP.2020.2990597},\n\tabstract = {We consider the problem of detecting abrupt changes in the distribution of a multi-dimensional time series, with limited computing power and memory. In this paper, we propose a new, simple method for model-free online change-point detection that relies only on fast and light recursive statistics, inspired by the classical Exponential Weighted Moving Average algorithm (EWMA). The proposed idea is to compute two EWMA statistics on the stream of data with different forgetting factors, and to compare them. By doing so, we show that we implicitly compare recent samples with older ones, without the need to explicitly store them. Additionally, we leverage Random Features (RFs) to efficiently use the Maximum Mean Discrepancy as a distance between distributions, furthermore exploiting recent optical hardware to compute high-dimensional RFs in near constant time. We show that our method is significantly faster than usual non-parametric methods for a given accuracy.},\n\tjournal = {IEEE Transactions on Signal Processing},\n\tauthor = {Keriven, Nicolas and Garreau, Damien and Poli, Iacopo},\n\tyear = {2020},\n\tnote = {Conference Name: IEEE Transactions on Signal Processing},\n\tkeywords = {Brain modeling, Change detection algorithms, Computational modeling, Hilbert space, Kernel, Microsoft Windows, Radio frequency, Signal processing algorithms, Streaming media, method of moments, optical computing},\n\tpages = {3515--3528},\n}\n\n","author_short":["Keriven, N.","Garreau, D.","Poli, I."],"key":"keriven_newma_2020","id":"keriven_newma_2020","bibbaseid":"keriven-garreau-poli-newmaanewmethodforscalablemodelfreeonlinechangepointdetection-2020","role":"author","urls":{},"keyword":["Brain modeling","Change detection algorithms","Computational modeling","Hilbert space","Kernel","Microsoft Windows","Radio frequency","Signal processing algorithms","Streaming media","method of moments","optical computing"],"metadata":{"authorlinks":{}}},"bibtype":"article","biburl":"https://bibbase.org/zotero/mh_lenguyen","dataSources":["yanwtMpCcFaHzRwWb","iwKepCrWBps7ojhDx"],"keywords":["brain modeling","change detection algorithms","computational modeling","hilbert space","kernel","microsoft windows","radio frequency","signal processing algorithms","streaming media","method of moments","optical computing"],"search_terms":["newma","new","method","scalable","model","free","online","change","point","detection","keriven","garreau","poli"],"title":"NEWMA: A New Method for Scalable Model-Free Online Change-Point Detection","year":2020}