{"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/subject-cross-validation-in-human-activity","title":"Subject Cross Validation in Human Activity Recognition","arxiv_id":"1904.02666","date":"2019-04-04","proceeding":null,"authors":["Akbar Dehghani","Tristan Glatard","Emad Shihab"],"abstract":"K-fold Cross Validation is commonly used to evaluate classifiers and tune\ntheir hyperparameters. However, it assumes that data points are Independent and\nIdentically Distributed (i.i.d.) so that samples used in the training and test\nsets can be selected randomly and uniformly. In Human Activity Recognition\ndatasets, we note that the samples produced by the same subjects are likely to\nbe correlated due to diverse factors. Hence, k-fold cross validation may\noverestimate the performance of activity recognizers, in particular when\noverlapping sliding windows are used. In this paper, we investigate the effect\nof Subject Cross Validation on the performance of Human Activity Recognition,\nboth with non-overlapping and with overlapping sliding windows. Results show\nthat k-fold cross validation artificially increases the performance of\nrecognizers by about 10%, and even by 16% when overlapping windows are used. In\naddition, we do not observe any performance gain from the use of overlapping\nwindows. We conclude that Human Activity Recognition systems should be\nevaluated by Subject Cross Validation, and that overlapping windows are not\nworth their extra computational cost.","url_abs":"http://arxiv.org/abs/1904.02666v2","url_pdf":"http://arxiv.org/pdf/1904.02666v2.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":"subject-cross-validation-in-human-activity","repo_url":"https://github.com/big-data-lab-team/paper-generalizability-window-size","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"activity-recognition","task_name":"Activity Recognition"},{"task_slug":"human-activity-recognition","task_name":"Human Activity Recognition"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}