{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2024,8,5]],"date-time":"2024-08-05T10:41:23Z","timestamp":1722854483579},"reference-count":25,"publisher":"MDPI AG","issue":"3","license":[{"start":{"date-parts":[[2022,9,10]],"date-time":"2022-09-10T00:00:00Z","timestamp":1662768000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["MAKE"],"abstract":"Factorizable joint shift (FJS) was recently proposed as a type of dataset shift for which the complete characteristics can be estimated from feature data observations on the test dataset by a method called Joint Importance Aligning. For the multinomial (multiclass) classification setting, we derive a representation of factorizable joint shift in terms of the source (training) distribution, the target (test) prior class probabilities and the target marginal distribution of the features. On the basis of this result, we propose alternatives to joint importance aligning and, at the same time, point out that factorizable joint shift is not fully identifiable if no class label information on the test dataset is available and no additional assumptions are made. Other results of the paper include correction formulae for the posterior class probabilities both under general dataset shift and factorizable joint shift. In addition, we investigate the consequences of assuming factorizable joint shift for the bias caused by sample selection.<\/jats:p>","DOI":"10.3390\/make4030038","type":"journal-article","created":{"date-parts":[[2022,9,13]],"date-time":"2022-09-13T05:44:03Z","timestamp":1663047843000},"page":"779-802","source":"Crossref","is-referenced-by-count":0,"title":["Factorizable Joint Shift in Multinomial Classification"],"prefix":"10.3390","volume":"4","author":[{"ORCID":"http:\/\/orcid.org\/0000-0002-2750-2970","authenticated-orcid":false,"given":"Dirk","family":"Tasche","sequence":"first","affiliation":[{"name":"Independent Researcher, 8032 Zurich, Switzerland"}]}],"member":"1968","published-online":{"date-parts":[[2022,9,10]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","unstructured":"Qui\u00f1onero-Candela, J., Sugiyama, M., Schwaighofer, A., and Lawrence, N. (2009). When Training and Test Sets Are Different: Characterizing Learning Transfer. Dataset Shift in Machine Learning, The MIT Press. Chapter 1.","DOI":"10.7551\/mitpress\/9780262170055.001.0001"},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"521","DOI":"10.1016\/j.patcog.2011.06.019","article-title":"A unifying view on dataset shift in classification","volume":"45","author":"Raeder","year":"2012","journal-title":"Pattern Recognit."},{"key":"ref_3","unstructured":"He, H., Yang, Y., and Wang, H. (2021). Domain Adaptation with Factorizable Joint Shift. arXiv."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"21","DOI":"10.1162\/089976602753284446","article-title":"Adjusting the Outputs of a Classifier to New a Priori Probabilities: A Simple Procedure","volume":"14","author":"Saerens","year":"2001","journal-title":"Neural Comput."},{"key":"ref_5","unstructured":"Nebel, B. (2001, January 4\u201310). The Foundations of Cost-Sensitive Learning. Proceedings of the Seventeenth International Joint Conference on Artificial Intelligence, IJCAI 2001, Seattle, WA, USA."},{"key":"ref_6","unstructured":"Billingsley, P. (1986). Probability and Measure, John Wiley & Sons. [2nd ed.]."},{"key":"ref_7","doi-asserted-by":"crossref","unstructured":"Klenke, A. (2013). Probability Theory: A Comprehensive Course, Springer Science & Business Media.","DOI":"10.1007\/978-1-4471-5361-0"},{"key":"ref_8","unstructured":"Tasche, D. (2022). Class Prior Estimation under Covariate Shift: No Problem?. arXiv."},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"595","DOI":"10.1214\/13-AOAS709","article-title":"The role of the information set for forecasting\u2014With applications to risk management","volume":"8","author":"Holzmann","year":"2014","journal-title":"Ann. Appl. Stat."},{"key":"ref_10","unstructured":"Chaudhuri, K., and Sugiyama, M. (2019, January 16\u201318). Support and Invertibility in Domain-Invariant Representations. Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics, Okinawa, Japan."},{"key":"ref_11","unstructured":"Scott, C. (2019, January 22\u201324). A Generalized Neyman-Pearson Criterion for Optimal Domain Adaptation. Proceedings of the Machine Learning Research, 30th International Conference on Algorithmic Learning Theory, Chicago, IL, USA."},{"key":"ref_12","first-page":"74:1","article-title":"A Review on Quantification Learning","volume":"50","author":"Chawla","year":"2017","journal-title":"ACM Comput. Surv."},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"63","DOI":"10.21314\/JCR.2013.169","article-title":"The art of probability-of-default curve calibration","volume":"9","author":"Tasche","year":"2013","journal-title":"J. Credit. Risk"},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"110","DOI":"10.1016\/j.neunet.2013.11.010","article-title":"Semi-supervised learning of class balance under class-prior change by distribution matching","volume":"50","author":"Sugiyama","year":"2014","journal-title":"Neural Netw."},{"key":"ref_15","unstructured":"Tasche, D. (2013). The Law of Total Odds. arXiv."},{"key":"ref_16","unstructured":"Zhang, K., Sch\u00f6lkopf, B., Muandet, K., and Wang, Z. (2013, January 17\u201319). Domain Adaptation Under Target and Conditional Shift. Proceedings of the 30th International Conference on International Conference on Machine Learning\u2014Volume 28, ICML\u201913, Atlanta, GA, USA."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"377","DOI":"10.1016\/j.csda.2012.07.007","article-title":"Drift mining in data: A framework for addressing drift in classification","volume":"57","author":"Hofer","year":"2013","journal-title":"Comput. Stat. Data Anal."},{"key":"ref_18","unstructured":"Dy, J., and Krause, A. (2018, January 10\u201315). Detecting and Correcting for Label Shift with Black Box Predictors. Proceedings of the 35th International Conference on Machine Learning, Stockholm, Sweden."},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"1356","DOI":"10.1080\/02331888.2021.2016767","article-title":"Calibrating sufficiently","volume":"55","author":"Tasche","year":"2021","journal-title":"Statistics"},{"key":"ref_20","doi-asserted-by":"crossref","unstructured":"Devroye, L., Gy\u00f6rfi, L., and Lugosi, G. (1996). A Probabilistic Theory of Pattern Recognition, Springer.","DOI":"10.1007\/978-1-4612-0711-5"},{"key":"ref_21","first-page":"19276","article-title":"Domain Adaptation with Conditional Distribution Matching and Generalized Label Shift","volume":"Volume 33","author":"Larochelle","year":"2020","journal-title":"Proceedings of the Advances in Neural Information Processing Systems"},{"key":"ref_22","first-page":"1","article-title":"Fisher Consistency for Prior Probability Shift","volume":"18","author":"Tasche","year":"2017","journal-title":"J. Mach. Learn. Res."},{"key":"ref_23","doi-asserted-by":"crossref","unstructured":"Qui\u00f1onero-Candela, J., Sugiyama, M., Schwaighofer, A., and Lawrence, N. (2009). Binary Classification under Sample Selection Bias. Dataset Shift in Machine Learning, The MIT Press. Chapter 3.","DOI":"10.7551\/mitpress\/9780262170055.001.0001"},{"key":"ref_24","doi-asserted-by":"crossref","unstructured":"Zadrozny, B. (2004, January 4\u20138). Learning and Evaluating Classifiers under Sample Selection Bias. Proceedings of the Twenty-First International Conference on Machine Learning, ICML\u201904, Banff, AB, Canada.","DOI":"10.1145\/1015330.1015425"},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Klebaner, F. (2005). Introduction to Stochastic Calculus with Applications, Imperial College Press. [2nd ed.].","DOI":"10.1142\/p386"}],"container-title":["Machine Learning and Knowledge Extraction"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2504-4990\/4\/3\/38\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,8,5]],"date-time":"2024-08-05T09:43:14Z","timestamp":1722850994000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2504-4990\/4\/3\/38"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,9,10]]},"references-count":25,"journal-issue":{"issue":"3","published-online":{"date-parts":[[2022,9]]}},"alternative-id":["make4030038"],"URL":"https:\/\/doi.org\/10.3390\/make4030038","relation":{},"ISSN":["2504-4990"],"issn-type":[{"type":"electronic","value":"2504-4990"}],"subject":[],"published":{"date-parts":[[2022,9,10]]}}}