{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/changepoint-detection-in-the-presence-of","title":"Changepoint Detection in the Presence of Outliers","arxiv_id":"1609.07363","date":"2016-09-23","proceeding":null,"authors":["Paul Fearnhead","Guillem Rigaill"],"abstract":"Many traditional methods for identifying changepoints can struggle in the\npresence of outliers, or when the noise is heavy-tailed. Often they will infer\nadditional changepoints in order to fit the outliers. To overcome this problem,\ndata often needs to be pre-processed to remove outliers, though this is\ndifficult for applications where the data needs to be analysed online. We\npresent an approach to changepoint detection that is robust to the presence of\noutliers. The idea is to adapt existing penalised cost approaches for detecting\nchanges so that they use loss functions that are less sensitive to outliers. We\nargue that loss functions that are bounded, such as the classical biweight\nloss, are particularly suitable -- as we show that only bounded loss functions\nare robust to arbitrarily extreme outliers. We present an efficient dynamic\nprogramming algorithm that can find the optimal segmentation under our\npenalised cost criteria. Importantly, this algorithm can be used in settings\nwhere the data needs to be analysed online. We show that we can consistently\nestimate the number of changepoints, and accurately estimate their locations,\nusing the biweight loss function. We demonstrate the usefulness of our approach\nfor applications such as analysing well-log data, detecting copy number\nvariation, and detecting tampering of wireless devices.","url_abs":"http://arxiv.org/abs/1609.07363v2","url_pdf":"http://arxiv.org/pdf/1609.07363v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"changepoint-detection-in-the-presence-of","repo_url":"https://github.com/guillemr/robust-fpop","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}