{"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/non-normal-mixtures-of-experts","title":"Non-Normal Mixtures of Experts","arxiv_id":"1506.06707","date":"2015-06-22","proceeding":null,"authors":["Faicel Chamroukhi"],"abstract":"Mixture of Experts (MoE) is a popular framework for modeling heterogeneity in\ndata for regression, classification and clustering. For continuous data which\nwe consider here in the context of regression and cluster analysis, MoE usually\nuse normal experts, that is, expert components following the Gaussian\ndistribution. However, for a set of data containing a group or groups of\nobservations with asymmetric behavior, heavy tails or atypical observations,\nthe use of normal experts may be unsuitable and can unduly affect the fit of\nthe MoE model. In this paper, we introduce new non-normal mixture of experts\n(NNMoE) which can deal with these issues regarding possibly skewed,\nheavy-tailed data and with outliers. The proposed models are the skew-normal\nMoE and the robust $t$ MoE and skew $t$ MoE, respectively named SNMoE, TMoE and\nSTMoE. We develop dedicated expectation-maximization (EM) and expectation\nconditional maximization (ECM) algorithms to estimate the parameters of the\nproposed models by monotonically maximizing the observed data log-likelihood.\nWe describe how the presented models can be used in prediction and in\nmodel-based clustering of regression data. Numerical experiments carried out on\nsimulated data show the effectiveness and the robustness of the proposed models\nin terms modeling non-linear regression functions as well as in model-based\nclustering. Then, to show their usefulness for practical applications, the\nproposed models are applied to the real-world data of tone perception for\nmusical data analysis, and the one of temperature anomalies for the analysis of\nclimate change data.","url_abs":"http://arxiv.org/abs/1506.06707v2","url_pdf":"http://arxiv.org/pdf/1506.06707v2.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":"non-normal-mixtures-of-experts","repo_url":"https://github.com/cran/meteorits","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null},{"paper_slug":"non-normal-mixtures-of-experts","repo_url":"https://github.com/fchamroukhi/MEteorits","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null},{"paper_slug":"non-normal-mixtures-of-experts","repo_url":"https://github.com/fchamroukhi/SNMoE","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"clustering","task_name":"Clustering"},{"task_slug":"mixture-of-experts","task_name":"Mixture-of-Experts"},{"task_slug":"regression-1","task_name":"regression"}],"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}