{"id":23785,"date":"2021-01-15T09:26:23","date_gmt":"2021-01-15T08:26:23","guid":{"rendered":"https:\/\/datascientest.com\/?p=23785"},"modified":"2023-11-09T14:10:21","modified_gmt":"2023-11-09T14:10:21","slug":"svm","status":"publish","type":"post","link":"https:\/\/datascientest.com\/svm","title":{"rendered":"SVM, quoi, comment, pourquoi ?"},"content":{"rendered":"\t\t<div data-elementor-type=\"wp-post\" data-elementor-id=\"23785\" class=\"elementor elementor-23785\" data-elementor-post-type=\"post\">\n\t\t\t\t\t\t\t<div class=\"elementor-element elementor-element-88d4a1c e-flex e-con-boxed e-con\" data-id=\"88d4a1c\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-f67a110 elementor-widget elementor-widget-heading\" data-id=\"f67a110\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style>\/*! elementor - v3.15.0 - 02-08-2023 *\/\n.elementor-heading-title{padding:0;margin:0;line-height:1}.elementor-widget-heading .elementor-heading-title[class*=elementor-size-]>a{color:inherit;font-size:inherit;line-height:inherit}.elementor-widget-heading .elementor-heading-title.elementor-size-small{font-size:15px}.elementor-widget-heading .elementor-heading-title.elementor-size-medium{font-size:19px}.elementor-widget-heading .elementor-heading-title.elementor-size-large{font-size:29px}.elementor-widget-heading .elementor-heading-title.elementor-size-xl{font-size:39px}.elementor-widget-heading .elementor-heading-title.elementor-size-xxl{font-size:59px}<\/style><h2 class=\"elementor-heading-title elementor-size-default\">Bienvenue dans la premi\u00e8re partie de cette mini-s\u00e9rie d\u00e9di\u00e9e aux SVM. <\/h2>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-2a035a4 e-flex e-con-boxed e-con\" data-id=\"2a035a4\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-e207e62 elementor-widget elementor-widget-text-editor\" data-id=\"e207e62\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style>\/*! elementor - v3.15.0 - 02-08-2023 *\/\n.elementor-widget-text-editor.elementor-drop-cap-view-stacked .elementor-drop-cap{background-color:#69727d;color:#fff}.elementor-widget-text-editor.elementor-drop-cap-view-framed .elementor-drop-cap{color:#69727d;border:3px solid;background-color:transparent}.elementor-widget-text-editor:not(.elementor-drop-cap-view-default) .elementor-drop-cap{margin-top:8px}.elementor-widget-text-editor:not(.elementor-drop-cap-view-default) .elementor-drop-cap-letter{width:1em;height:1em}.elementor-widget-text-editor .elementor-drop-cap{float:left;text-align:center;line-height:1;font-size:50px}.elementor-widget-text-editor .elementor-drop-cap-letter{display:inline-block}<\/style>\t\t\t\t<p><span style=\"font-weight: 400;\">Quand on pense <a href=\"https:\/\/datascientest.com\/machine-learning-tout-savoir\">Machine Learning<\/a>, on pense g\u00e9n\u00e9ralement \u00e0 des m\u00e9thodes r\u00e9centes, encore mal comprises, et utilis\u00e9es empiriquement, \u00e0 la mani\u00e8re d\u2019un apprenti sorcier.\u00a0<\/span><\/p><p><span style=\"font-weight: 400;\">En effet, que ce soit les algorithmes g\u00e9n\u00e9tiques, les <\/span><a href=\"https:\/\/datascientest.com\/fonctionnement-des-reseaux-neurones\" target=\"_blank\" rel=\"noopener\">r\u00e9seaux de neurones<\/a><span style=\"font-weight: 400;\">, le<\/span><a href=\"https:\/\/datascientest.com\/algorithmes-de-boosting-adaboost-gradient-boosting-xgboost\" target=\"_blank\" rel=\"noopener\"> Boosting<\/a><span style=\"font-weight: 400;\">&#8230; on qualifie souvent ces algorithmes de \u00ab\u00a0bo\u00eetes noires\u00a0\u00bb. Autrement dit, m\u00eame s\u2019il est facile de commenter l\u2019entr\u00e9e et la sortie de l\u2019algorithme ind\u00e9pendamment, lorsqu\u2019il s\u2019agit de d\u00e9tailler la contribution exacte de chaque entr\u00e9e au r\u00e9sultat final, c\u2019est une autre paire de manches. La faute \u00e0 un trop grand nombre de param\u00e8tres, \u00e0 l\u2019utilisation d\u2019\u00e9v\u00e9nements al\u00e9atoires&#8230;<\/span><\/p><p><span style=\"font-weight: 400;\">N\u00e9anmoins, parmi cette myriade d\u2019algorithmes tous plus complexes les uns que les autres, s\u2019en trouve un, dont l\u2019\u00e2ge avanc\u00e9 d\u00e9note avec son efficacit\u00e9 et sa simplicit\u00e9.<\/span><\/p><p><span style=\"font-weight: 400;\">Conceptualis\u00e9 par Vladimir Vapnik d\u00e8s le d\u00e9but des ann\u00e9es 60, il fait son entr\u00e9e dans le milieu de l\u2019informatique en 1990. D\u00e8s lors, il devient une r\u00e9f\u00e9rence pour tous ceux qui s\u2019int\u00e9ressent de pr\u00e8s ou de loin aux <\/span><b>techniques d\u2019apprentissage supervis\u00e9. <\/b><span style=\"font-weight: 400;\">Encore aujourd\u2019hui, il continue d\u2019\u00eatre enseign\u00e9, \u00e9tudi\u00e9 et massivement utilis\u00e9. Vous l\u2019aurez compris, nous allons parler ici des <\/span><b>Machines \u00e0 Vecteurs de Support, aussi appel\u00e9 SVM<\/b><span style=\"font-weight: 400;\"> pour <\/span><b>Support Vector Machine<\/b><span style=\"font-weight: 400;\">.\u00a0<\/span><\/p><p><span style=\"font-weight: 400;\">Warning : Parler de la simplicit\u00e9 d\u2019un tel algorithme revient \u00e0 parler de la petitesse de la Lune. On sait qu\u2019elle est petite \u00e0 c\u00f4t\u00e9 de ce qui l\u2019entoure, mais on pourrait quand m\u00eame pas en faire le tour \u00e0 pied.\u00a0<\/span><\/p><p><span style=\"font-weight: 400;\">Cet article sera divis\u00e9 en 2 parties. Dans celle-ci, nous allons vous pr\u00e9senter le <strong>principe de base des Margin Classifiers<\/strong>, et les limites de ces derniers.<\/span><\/p><p><span style=\"font-weight: 400;\">Dans la seconde, nous rentrerons davantage dans le d\u00e9tail des SVM, l\u2019objectif \u00e9tant de donner une vue d\u2019ensemble sur leur fonctionnement<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-04488e7 e-flex e-con-boxed e-con\" data-id=\"04488e7\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-ad0b37d elementor-widget elementor-widget-heading\" data-id=\"ad0b37d\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Maximal Margin Classifiers<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-3a41900 elementor-widget elementor-widget-text-editor\" data-id=\"3a41900\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Comme souvent en Sciences, avant de s\u2019attaquer au probl\u00e8me en tant que tel, on pr\u00e9f\u00e8re <\/span><b>\u00e9tudier un cas particulier<\/b><span style=\"font-weight: 400;\">, plus simple \u00e0 r\u00e9soudre. Nous allons donc commencer cette pr\u00e9sentation par l\u2019\u00e9tude d\u2019un cousin des SVM,<\/span><b> les Classifieurs \u00e0 Marge Maximal<\/b><span style=\"font-weight: 400;\">.\u00a0<\/span><\/p><p><span style=\"font-weight: 400;\">Supposons que nous ayons un jeu de donn\u00e9es contenant 2 types d\u2019observations : les \u00ab\u00a0+\u00a0\u00bb et les \u00ab\u00a0-\u00ab\u00a0. Notre objectif est de pouvoir <\/span><b>pr\u00e9dire la classe<\/b><span style=\"font-weight: 400;\"> (\u00ab\u00a0+\u00a0\u00bb ou \u00ab\u00a0-\u00ab\u00a0) d\u2019une nouvelle observation, ici repr\u00e9sent\u00e9e par un \u00ab\u00a0x\u00a0\u00bb. <\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-4f911d0 elementor-widget elementor-widget-image\" data-id=\"4f911d0\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style>\/*! elementor - v3.15.0 - 02-08-2023 *\/\n.elementor-widget-image{text-align:center}.elementor-widget-image a{display:inline-block}.elementor-widget-image a img[src$=\".svg\"]{width:48px}.elementor-widget-image img{vertical-align:middle;display:inline-block}<\/style>\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"349\" height=\"231\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/12\/svm.png\" class=\"attachment-full size-full wp-image-23788\" alt=\"svm\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/12\/svm.png 349w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/12\/svm-300x199.png 300w\" sizes=\"(max-width: 349px) 100vw, 349px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-5acbaea elementor-widget elementor-widget-text-editor\" data-id=\"5acbaea\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<blockquote><p><span style=\"font-weight: 400;\">Instinctivement, on comprend bien que l\u2019observation devrait \u00eatre class\u00e9e parmi les \u00ab\u00a0&#8211;\u00a0\u00bb ici. Mais il est difficile d\u2019expliquer les raisons qui nous permettent d\u2019affirmer cela.<\/span><\/p><\/blockquote><p><span style=\"font-weight: 400;\">La m\u00e9thode que l\u2019on va appliquer est la suivante : <\/span><\/p><p><b>Tracer une droite<\/b><span style=\"font-weight: 400;\"> s\u00e9parant les 2 classes. Si l\u2019observation se trouve au-dessus de cette droite, on dira qu\u2019il s\u2019agit d\u2019un \u00ab\u00a0+\u00a0\u00bb, sinon d\u2019un \u00ab\u00a0-\u00ab\u00a0. Ici par exemple, l\u2019observation \u00ab\u00a0x\u00a0\u00bb se trouve en dessus de la droite, il s\u2019agit donc d\u2019un \u00ab\u00a0-\u00ab\u00a0.\u00a0<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-1722478 elementor-widget elementor-widget-image\" data-id=\"1722478\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"349\" height=\"231\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/12\/svm2.png\" class=\"attachment-full size-full wp-image-23790\" alt=\"\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/12\/svm2.png 349w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/12\/svm2-300x199.png 300w\" sizes=\"(max-width: 349px) 100vw, 349px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-52115ea elementor-widget elementor-widget-text-editor\" data-id=\"52115ea\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">La m\u00e9thode semble fonctionner correctement. Cependant, dans le cas particulier de notre probl\u00e8me, <\/span><b>il existe une infinit\u00e9 de droites s\u00e9parant parfaitement les 2 ensembles<\/b><span style=\"font-weight: 400;\">. Chacune d\u2019elle d\u00e9limite des espaces diff\u00e9rents et constitue une r\u00e9ponse diff\u00e9rente au probl\u00e8me. La question est alors de savoir comment choisir <\/span><b>la meilleure droite<\/b><span style=\"font-weight: 400;\"> parmi toutes les droites possibles.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-9566828 elementor-widget elementor-widget-image\" data-id=\"9566828\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"349\" height=\"231\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/12\/svm3.png\" class=\"attachment-full size-full wp-image-23798\" alt=\"\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/12\/svm3.png 349w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/12\/svm3-300x199.png 300w\" sizes=\"(max-width: 349px) 100vw, 349px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-178457b elementor-widget elementor-widget-text-editor\" data-id=\"178457b\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Sur la figure pr\u00e9c\u00e9dente, il semble logique de dire que la droite centrale s\u00e9pare mieux les donn\u00e9es que les 2 autres. En effet, comme elle est \u00e9loign\u00e9e \u00e0 la fois des \u00ab\u00a0+\u00a0\u00bb et des \u00ab\u00a0-\u00ab\u00a0, elle ne risque pas de classer une observation dans une classe alors qu\u2019elle se trouve en r\u00e9alit\u00e9 tr\u00e8s proche de l\u2019autre classe.\u00a0<\/span><\/p><blockquote><p><span style=\"font-weight: 400;\">En math\u00e9matiques, la distance qui s\u00e9pare une droite de l\u2019observation la plus proche est appel\u00e9e <\/span><b>la marge<\/b><span style=\"font-weight: 400;\">. <\/span><span style=\"font-weight: 400;\">La droite centrale a une marge plus importante que les deux autres.\u00a0<\/span><\/p><\/blockquote><p><span style=\"font-weight: 400;\">Autrement dit, la droite qui s\u00e9pare le mieux les 2 classes est celle dont la <\/span><b>marge est la plus \u00e9lev\u00e9e. <\/b><span style=\"font-weight: 400;\">Vous comprenez d\u00e9sormais d\u2019o\u00f9 vient le terme de \u00ab\u00a0Maximal Margin Classifiers\u00a0\u00bb ?<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-748bf81 elementor-widget elementor-widget-image\" data-id=\"748bf81\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"349\" height=\"231\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/12\/svm4.png\" class=\"attachment-full size-full wp-image-23802\" alt=\"\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/12\/svm4.png 349w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/12\/svm4-300x199.png 300w\" sizes=\"(max-width: 349px) 100vw, 349px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-faa050b elementor-widget elementor-widget-text-editor\" data-id=\"faa050b\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Dans le cas de notre exemple, la droite continue est la meilleure droite de s\u00e9paration. La marge, qui est repr\u00e9sent\u00e9e en pointill\u00e9, est maximale. Vous aurez remarqu\u00e9 que la m\u00eame distance s\u00e9pare la droite des 3 observations les plus proches. C\u2019est une caract\u00e9ristique que l\u2019on retrouve syst\u00e9matiquement pour la meilleure droite.<\/span><\/p><p><span style=\"font-weight: 400;\">Il existe une m\u00e9thode permettant de la calculer \u00e0 tous les coups, l\u2019algorithme Maximal Margin Classifier utilise cette m\u00e9thode pour\u00a0<\/span><\/p><blockquote><p><span style=\"font-weight: 400;\">Ce qu\u2019il faut retenir de cette introduction, c\u2019est qu\u2019il suffit de trouver la droite avec la marge la plus grande pour pouvoir classer toutes les nouvelles observations<\/span><span style=\"font-weight: 400;\">.<\/span><\/p><\/blockquote><p><span style=\"font-weight: 400;\"> De plus, lorsqu\u2019elle existe, il est relativement simple de calculer l\u2019\u00e9quation de la droite de s\u00e9paration (si le d\u00e9tail des calculs vous int\u00e9resse, je vous recommande la<\/span> <a href=\"https:\/\/www.youtube.com\/watch?v=_PwhiWxHK8o\" target=\"_blank\" rel=\"noopener\">vid\u00e9o\u00a0<\/a><span style=\"font-weight: 400;\"><a href=\"https:\/\/www.youtube.com\/watch?v=_PwhiWxHK8o)\" target=\"_blank\" rel=\"noopener\">d<\/a><a href=\"https:\/\/www.youtube.com\/watch?v=_PwhiWxHK8o)\" target=\"_blank\" rel=\"noopener\">e<\/a><a href=\"https:\/\/www.youtube.com\/watch?v=_PwhiWxHK8o)\" target=\"_blank\" rel=\"noopener\"> MIT<\/a><a href=\"https:\/\/www.youtube.com\/watch?v=_PwhiWxHK8o)\" target=\"_blank\" rel=\"noopener\"> OpenCourseWare<\/a> qui traite pr\u00e9cis\u00e9ment de ce sujet .\u00a0<\/span><\/p><p><span style=\"font-weight: 400;\">On tient alors une m\u00e9thode <\/span><b>simple <\/b><span style=\"font-weight: 400;\">et <\/span><b>rapide<\/b><span style=\"font-weight: 400;\"> pour r\u00e9pondre au probl\u00e8me pos\u00e9.\u00a0<\/span><\/p><p><span style=\"font-weight: 400;\">Cependant, la m\u00e9thode que nous venons de voir a deux gros d\u00e9fauts. En effet, elle est tr\u00e8s<\/span><b> sensible aux valeurs extr\u00eames<\/b><span style=\"font-weight: 400;\"> (aussi appel\u00e9es outliers).<\/span><\/p><p><span style=\"font-weight: 400; color: #000000;\">L\u2019arriv\u00e9e d\u2019un outlier (le \u00ab\u00a0&#8211;\u00a0\u00bb le plus haut) perturbe compl\u00e8tement le r\u00e9sultat.<\/span><\/p><p><span style=\"font-weight: 400; color: #000000;\">Pire encore, au cas o\u00f9 vous vous seriez pos\u00e9 la question : Non, il n\u2019existe pas toujours de droite s\u00e9parant parfaitement les 2 classes. C\u2019est par exemple le cas dans l\u2019image qui suit :<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-bf6c823 elementor-widget elementor-widget-image\" data-id=\"bf6c823\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"349\" height=\"231\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/12\/svm6.png\" class=\"attachment-full size-full wp-image-23804\" alt=\"\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/12\/svm6.png 349w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/12\/svm6-300x199.png 300w\" sizes=\"(max-width: 349px) 100vw, 349px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-fffab49 elementor-widget elementor-widget-text-editor\" data-id=\"fffab49\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Nous nous retrouvons finalement avec un algorithme qui, dans certains cas, <\/span><b>ne fonctionne pas du tout<\/b><span style=\"font-weight: 400;\">.<\/span><\/p><p><span style=\"color: #000000;\"><span style=\"font-weight: 400;\">Dans le prochain article d\u00e9di\u00e9 aux SVM, nous verrons comment donner davantage de flexibilit\u00e9 \u00e0 notre algorithme, en l\u2019autorisant \u00e0 commettre quelques erreurs de classification, pour am\u00e9liorer ses performances globales.<\/span><\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-3a674bd e-flex e-con-boxed e-con\" data-id=\"3a674bd\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-f549cb8 elementor-widget elementor-widget-text-editor\" data-id=\"f549cb8\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>Les <a href=\"https:\/\/datascientest.com\/blog-data-ia-actualites\/machine-learning\">techniques de Machine Learning<\/a> vous int\u00e9ressent ? Vous souhaitez apprendre \u00e0 les ma\u00eetriser ? Nos <strong>formations en data science<\/strong> sont faites pour \u00e7a !<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-9e72ea4 elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"9e72ea4\" data-element_type=\"widget\" data-widget_type=\"button.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<div class=\"elementor-button-wrapper\">\n\t\t\t<a class=\"elementor-button elementor-button-link elementor-size-sm elementor-animation-grow\" href=\"https:\/\/datascientest.com\/formations-data\">\n\t\t\t\t\t\t<span class=\"elementor-button-content-wrapper\">\n\t\t\t\t\t\t<span class=\"elementor-button-text\">D\u00e9couvrir nos formations<\/span>\n\t\t<\/span>\n\t\t\t\t\t<\/a>\n\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t\t\t<\/div>\n\t\t","protected":false},"excerpt":{"rendered":"<p>Bienvenue dans la premi\u00e8re partie de cette mini-s\u00e9rie d\u00e9di\u00e9e aux SVM. Quand on pense Machine Learning, on pense g\u00e9n\u00e9ralement \u00e0 des m\u00e9thodes r\u00e9centes, encore mal comprises, et utilis\u00e9es empiriquement, \u00e0 la mani\u00e8re d\u2019un apprenti sorcier.&nbsp; En effet, que ce soit les algorithmes g\u00e9n\u00e9tiques, les r\u00e9seaux de neurones, le Boosting&#8230; on qualifie souvent ces algorithmes de [&hellip;]<\/p>\n","protected":false},"author":77,"featured_media":26183,"comment_status":"open","ping_status":"open","sticky":false,"template":"elementor_theme","format":"standard","meta":{"content-type":"","inline_featured_image":false},"categories":[2497],"tags":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v21.7 (Yoast SEO v21.7) - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>SVM: D\u00e9finition, application et exemples- multi.datascientest.com<\/title>\n<meta name=\"description\" content=\"D\u00e9couvrez le support vector machine ou SVM, une technique d&#039;apprentissage supervis\u00e9 qui permet de r\u00e9soudre les probl\u00e8mes de r\u00e9gression\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/datascientest.com\/svm\" \/>\n<meta property=\"og:locale\" content=\"fr_FR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"SVM, quoi, comment, pourquoi ?\" \/>\n<meta property=\"og:url\" content=\"https:\/\/datascientest.com\/svm\" \/>\n<meta property=\"og:site_name\" content=\"Formation Data Science | DataScientest.com\" \/>\n<meta property=\"article:published_time\" content=\"2021-01-15T08:26:23+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2023-11-09T14:10:21+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/12\/illu_SVM_blog-42.png\" \/>\n\t<meta property=\"og:image:width\" content=\"2560\" \/>\n\t<meta property=\"og:image:height\" content=\"1406\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"Raphael Kassel\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u00c9crit par\" \/>\n\t<meta name=\"twitter:data1\" content=\"Raphael Kassel\" \/>\n\t<meta name=\"twitter:label2\" content=\"Dur\u00e9e de lecture estim\u00e9e\" \/>\n\t<meta name=\"twitter:data2\" content=\"6 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/datascientest.com\/svm\",\"url\":\"https:\/\/datascientest.com\/svm\",\"name\":\"SVM: D\u00e9finition, application et exemples- multi.datascientest.com\",\"isPartOf\":{\"@id\":\"https:\/\/datascientest.com\/#website\"},\"datePublished\":\"2021-01-15T08:26:23+00:00\",\"dateModified\":\"2023-11-09T14:10:21+00:00\",\"author\":{\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/6eca16f63daaafcbc96e15549b9d4ee1\"},\"description\":\"D\u00e9couvrez le support vector machine ou SVM, une technique d'apprentissage supervis\u00e9 qui permet de r\u00e9soudre les probl\u00e8mes de r\u00e9gression\",\"inLanguage\":\"fr-FR\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/datascientest.com\/svm\"]}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/datascientest.com\/#website\",\"url\":\"https:\/\/datascientest.com\/\",\"name\":\"Formation Data Science | DataScientest.com\",\"description\":\"Leader de la formation en Data Science\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/datascientest.com\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"fr-FR\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/6eca16f63daaafcbc96e15549b9d4ee1\",\"name\":\"Raphael Kassel\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"fr-FR\",\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/raph.jpeg\",\"contentUrl\":\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/raph.jpeg\",\"caption\":\"Raphael Kassel\"},\"description\":\"Raphael Kassel est le Head of Data Programs chez DataScientest. Il supervise une \u00e9quipe Data, impl\u00e9mente de nouveaux parcours comme \\\"Chef de projet IA\\\" et \\\"Data Product Manager\\\", et contribue \u00e0 la modernisation des op\u00e9rations du groupe. En collaboration \u00e9troite avec l'\u00e9quipe de d\u00e9veloppement, il adapte l'accompagnement acad\u00e9mique aux \u00e9volutions de l'\u00e9quipe data et \u00e0 l'internationalisation.\",\"sameAs\":[\"https:\/\/www.linkedin.com\/in\/raphaelkassel\/\",\"Data Science\"],\"url\":\"https:\/\/datascientest.com\/author\/raphaelkassel\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"SVM: D\u00e9finition, application et exemples- multi.datascientest.com","description":"D\u00e9couvrez le support vector machine ou SVM, une technique d'apprentissage supervis\u00e9 qui permet de r\u00e9soudre les probl\u00e8mes de r\u00e9gression","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/datascientest.com\/svm","og_locale":"fr_FR","og_type":"article","og_title":"SVM, quoi, comment, pourquoi ?","og_url":"https:\/\/datascientest.com\/svm","og_site_name":"Formation Data Science | DataScientest.com","article_published_time":"2021-01-15T08:26:23+00:00","article_modified_time":"2023-11-09T14:10:21+00:00","og_image":[{"width":2560,"height":1406,"url":"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/12\/illu_SVM_blog-42.png","type":"image\/png"}],"author":"Raphael Kassel","twitter_card":"summary_large_image","twitter_misc":{"\u00c9crit par":"Raphael Kassel","Dur\u00e9e de lecture estim\u00e9e":"6 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/datascientest.com\/svm","url":"https:\/\/datascientest.com\/svm","name":"SVM: D\u00e9finition, application et exemples- multi.datascientest.com","isPartOf":{"@id":"https:\/\/datascientest.com\/#website"},"datePublished":"2021-01-15T08:26:23+00:00","dateModified":"2023-11-09T14:10:21+00:00","author":{"@id":"https:\/\/datascientest.com\/#\/schema\/person\/6eca16f63daaafcbc96e15549b9d4ee1"},"description":"D\u00e9couvrez le support vector machine ou SVM, une technique d'apprentissage supervis\u00e9 qui permet de r\u00e9soudre les probl\u00e8mes de r\u00e9gression","inLanguage":"fr-FR","potentialAction":[{"@type":"ReadAction","target":["https:\/\/datascientest.com\/svm"]}]},{"@type":"WebSite","@id":"https:\/\/datascientest.com\/#website","url":"https:\/\/datascientest.com\/","name":"Formation Data Science | DataScientest.com","description":"Leader de la formation en Data Science","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/datascientest.com\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"fr-FR"},{"@type":"Person","@id":"https:\/\/datascientest.com\/#\/schema\/person\/6eca16f63daaafcbc96e15549b9d4ee1","name":"Raphael Kassel","image":{"@type":"ImageObject","inLanguage":"fr-FR","@id":"https:\/\/datascientest.com\/#\/schema\/person\/image\/","url":"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/raph.jpeg","contentUrl":"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/raph.jpeg","caption":"Raphael Kassel"},"description":"Raphael Kassel est le Head of Data Programs chez DataScientest. Il supervise une \u00e9quipe Data, impl\u00e9mente de nouveaux parcours comme \"Chef de projet IA\" et \"Data Product Manager\", et contribue \u00e0 la modernisation des op\u00e9rations du groupe. En collaboration \u00e9troite avec l'\u00e9quipe de d\u00e9veloppement, il adapte l'accompagnement acad\u00e9mique aux \u00e9volutions de l'\u00e9quipe data et \u00e0 l'internationalisation.","sameAs":["https:\/\/www.linkedin.com\/in\/raphaelkassel\/","Data Science"],"url":"https:\/\/datascientest.com\/author\/raphaelkassel"}]}},"_links":{"self":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts\/23785"}],"collection":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/users\/77"}],"replies":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/comments?post=23785"}],"version-history":[{"count":0,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts\/23785\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/media\/26183"}],"wp:attachment":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/media?parent=23785"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/categories?post=23785"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/tags?post=23785"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}