{"id":259331,"date":"2025-01-14T06:30:00","date_gmt":"2025-01-14T06:30:00","guid":{"rendered":"https:\/\/datascientest.com\/?p=259331"},"modified":"2025-01-10T13:01:09","modified_gmt":"2025-01-10T13:01:09","slug":"fonction-de-cout-tout-savoir","status":"publish","type":"post","link":"https:\/\/datascientest.com\/fonction-de-cout-tout-savoir","title":{"rendered":"La fonction de co\u00fbt en IA : Tout ce qu&rsquo;il faut savoir"},"content":{"rendered":"\t\t<div data-elementor-type=\"wp-post\" data-elementor-id=\"259331\" class=\"elementor elementor-259331\" data-elementor-post-type=\"post\">\n\t\t\t\t\t\t\t<div class=\"elementor-element elementor-element-2a79321 e-flex e-con-boxed e-con\" data-id=\"2a79321\" 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-81112e1 elementor-widget elementor-widget-text-editor\" data-id=\"81112e1\" 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><strong>Si l\u2019erreur est humaine, elle n\u2019est pas exclusive \u00e0 l\u2019Homme. Les algorithmes d\u2019apprentissage automatique peuvent aussi faire des erreurs. Mais \u00e0 la diff\u00e9rence de nous, pauvres humains, il est possible de minimiser. Notamment en utilisant la fonction de co\u00fbt qui permet d\u2019\u00e9valuer la performance d\u2019un mod\u00e8le de machine learning.\u00a0<span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">Alors de quoi s\u2019agit-il exactement? Et comment l\u2019utiliser ? Data Scientest r\u00e9pond \u00e0 vos questions.\u00a0<\/span><\/strong><\/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-b8f56de e-flex e-con-boxed e-con\" data-id=\"b8f56de\" 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-f996689 elementor-widget elementor-widget-heading\" data-id=\"f996689\" 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><h3 class=\"elementor-heading-title elementor-size-default\">C\u2019est quoi la fonction de co\u00fbt ?<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-bd269cc elementor-widget elementor-widget-text-editor\" data-id=\"bd269cc\" 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;\">\u00c9galement appel\u00e9e fonction d\u2019erreur, fonction-objectif ou fonction de perte, la fonction de co\u00fbt <\/span><b>mesure l\u2019\u00e9cart entre les pr\u00e9dictions d\u2019un mod\u00e8le et les valeurs r\u00e9elles<\/b><span style=\"font-weight: 400;\">. Plus cet \u00e9cart est faible, plus le mod\u00e8le est performant. Cette formule math\u00e9matique joue ainsi un r\u00f4le de boussole ; elle indique si le mod\u00e8le est sur la bonne voie ou s\u2019il doit \u00eatre ajust\u00e9.<\/span><\/p><p><span style=\"font-weight: 400;\">Pour <a href=\"https:\/\/datascientest.com\/devenir-data-scientist-comment-faire\">les data scientists<\/a>, l\u2019objectif est simple : minimiser la fonction de co\u00fbt pour <\/span><a href=\"https:\/\/datascientest.com\/peut-on-avoir-confiance-aux-predictions-dune-ia-lapproche-lime\"><span style=\"font-weight: 400;\">am\u00e9liorer la pr\u00e9cision des pr\u00e9dictions.\u00a0<\/span><\/a><\/p><p><span style=\"font-weight: 400;\"><strong><span style=\"text-decoration: underline;\">Bon \u00e0 savoir<\/span><\/strong> : si la fonction de perte est tr\u00e8s utilis\u00e9e pour l\u2019entra\u00eenement des mod\u00e8les de machine learning, celle-ci a d\u2019autres applications. Par exemple, l\u2019industrie (pour minimiser les co\u00fbts de production), l\u2019\u00e9conomie (pour maximiser les rendements), la gestion des ressources humaines (pour choisir le bon candidat via une analyse co\u00fbt-b\u00e9n\u00e9fice), etc.\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-5fe402f elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"5fe402f\" 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\/formation-intelligence-artificielle\">\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\">Apprendre \u00e0 d\u00e9velopper un mod\u00e8le de ML<\/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<div class=\"elementor-element elementor-element-c4145e0 e-flex e-con-boxed e-con\" data-id=\"c4145e0\" 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-69b7fb2 elementor-widget elementor-widget-heading\" data-id=\"69b7fb2\" 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\">Comment utiliser la fonction de co\u00fbt ?<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-8a69105 elementor-widget elementor-widget-text-editor\" data-id=\"8a69105\" 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;\">Il n\u2019existe pas une fonction erreur, mais plusieurs. Chacune r\u00e9pondant \u00e0 un type de probl\u00e8me sp\u00e9cifique et aux caract\u00e9ristiques des donn\u00e9es du mod\u00e8le d\u2019apprentissage automatique. Voyons les principales.<\/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-fa34472 elementor-widget elementor-widget-heading\" data-id=\"fa34472\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h4 class=\"elementor-heading-title elementor-size-default\"><font size=\"4\">La moyenne des erreurs quadratique (Mean Squared Error - MSE)<\/font><\/h4>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-c3892ee elementor-widget elementor-widget-text-editor\" data-id=\"c3892ee\" 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;\">C\u2019est la fonction co\u00fbt la plus populaire en <\/span><a href=\"https:\/\/datascientest.com\/machine-learning-tout-savoir\"><span style=\"font-weight: 400;\">machine learning <\/span><\/a><span style=\"font-weight: 400;\">(notamment pour les probl\u00e8mes de r\u00e9gression). Cette m\u00e9thode des moindres carr\u00e9s p\u00e9nalise fortement les grandes erreurs en les \u00e9levant au carr\u00e9. Elle est particuli\u00e8rement <\/span><b>utile pour des pr\u00e9dictions o\u00f9 les \u00e9carts importants sont critiques.<\/b><\/p><p><span style=\"font-weight: 400;\">Alors comment l\u2019utiliser ?\u00a0<\/span><\/p><p><span style=\"font-weight: 400;\">Voici la formule math\u00e9matique : <\/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-1108d1c elementor-widget elementor-widget-image\" data-id=\"1108d1c\" 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=\"327\" height=\"84\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/12\/image2.png\" class=\"attachment-full size-full wp-image-259338\" alt=\"\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/12\/image2.png 327w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/12\/image2-300x77.png 300w\" sizes=\"(max-width: 327px) 100vw, 327px\" \/>\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-960343e elementor-widget elementor-widget-text-editor\" data-id=\"960343e\" 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;\">Et ses param\u00e8tres :\u00a0<\/span><\/p><ul><li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\"><em>J(\u03b8)<\/em>\u00a0: la valeur de la fonction de co\u00fbt (ou \u00ab\u00a0erreur moyenne quadratique\u00a0\u00bb) pour un ensemble donn\u00e9 de param\u00e8tres \u03b8<\/span><\/li><li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\"><em>m<\/em> : le nombre total d&rsquo;exemples dans le dataset.<\/span><\/li><li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\"><em>h<sub>\u03b8<\/sub>(x<sup>(i)<\/sup>)<\/em> : la pr\u00e9diction du mod\u00e8le pour l&rsquo;exemple i, souvent appel\u00e9e \u00ab\u00a0hypoth\u00e8se\u00a0\u00bb.<\/span><\/li><li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\"><em>y<sup>(i)<\/sup><\/em> : la valeur r\u00e9elle attendue pour l&rsquo;exemple i.<\/span><\/li><li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\"><em>(h<sub>\u03b8<\/sub>(x<sup>(i)<\/sup>)\u2212y<sup>(i)<\/sup>)<sup>2<\/sup><\/em> : le carr\u00e9 de l&rsquo;erreur entre la pr\u00e9diction et la vraie valeur pour l&rsquo;exemple i.<\/span><\/li><\/ul>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-fce5bf2 elementor-widget elementor-widget-image\" data-id=\"fce5bf2\" 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=\"1000\" height=\"571\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/12\/fonction-cout-machine-learning-datascientest-2.webp\" class=\"attachment-full size-full wp-image-259354\" alt=\"\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/12\/fonction-cout-machine-learning-datascientest-2.webp 1000w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/12\/fonction-cout-machine-learning-datascientest-2-300x171.webp 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/12\/fonction-cout-machine-learning-datascientest-2-768x439.webp 768w\" sizes=\"(max-width: 1000px) 100vw, 1000px\" \/>\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-a441963 elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"a441963\" 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\/formation-intelligence-artificielle\">\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\">Savoir utiliser la fonction de co\u00fbt en ML<\/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<div class=\"elementor-element elementor-element-9852482 elementor-widget elementor-widget-text-editor\" data-id=\"9852482\" 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;\">Par exemple, imaginons que l\u2019on souhaite entra\u00eener un mod\u00e8le pour pr\u00e9dire le prix d&rsquo;une maison en fonction de caract\u00e9ristiques comme la superficie, le nombre de chambres, l&rsquo;ann\u00e9e de construction, etc. Ici, le <a href=\"https:\/\/datascientest.com\/regression-lineaire-tout-savoir\">mod\u00e8le de r\u00e9gression lin\u00e9aire<\/a> permettra d\u2019\u00e9tablir une relation entre ces caract\u00e9ristiques (les entr\u00e9es = <em>x<\/em>) et le prix de la maison (la sortie attendue = <em>y<\/em>).<\/span><\/p><p><span style=\"font-weight: 400;\">\u00c0 partir de ces informations, il faudra r\u00e9aliser diff\u00e9rentes hypoth\u00e8ses avec <em>x<\/em> et <em>y<\/em>,\u00a0 jusqu\u2019\u00e0 trouver les mod\u00e8les pr\u00e9sentant la fonction co\u00fbt la plus faible possible. Autrement dit, r\u00e9duire les \u00e9carts pour que les pr\u00e9dictions soient les plus proches possibles des valeurs r\u00e9elles.<\/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-1dd6912 elementor-widget elementor-widget-heading\" data-id=\"1dd6912\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h4 class=\"elementor-heading-title elementor-size-default\"><font size=\"4\">L\u2019erreur absolue moyenne (Mean Absolute Error - MAE)<\/font><\/h4>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-6a5dafe elementor-widget elementor-widget-text-editor\" data-id=\"6a5dafe\" 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;\">Il s\u2019agit d\u2019une autre formule pour les mod\u00e8les de r\u00e9gression lin\u00e9aire. Mais \u00e0 la diff\u00e9rence de la pr\u00e9c\u00e9dente, cette fonction de co\u00fbt<\/span><b> mesure la moyenne des erreurs absolues entre les valeurs r\u00e9elles et les pr\u00e9dictions d\u2019un mod\u00e8le<\/b><span style=\"font-weight: 400;\">. Et comme la MAE n\u2019\u00e9l\u00e8ve pas les erreurs au carr\u00e9, elle est beaucoup<\/span><b> plus robuste face aux valeurs aberrantes.\u00a0<\/b><\/p><p><span style=\"font-weight: 400;\">Voici la formule math\u00e9matique :<\/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-030b307 elementor-widget elementor-widget-image\" data-id=\"030b307\" 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=\"288\" height=\"106\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/12\/image7.png\" class=\"attachment-full size-full wp-image-259339\" alt=\"\" loading=\"lazy\" \/>\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-aa07a87 elementor-widget elementor-widget-text-editor\" data-id=\"aa07a87\" 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<span style=\"font-weight: 400;\">Et ses param\u00e8tres :\u00a0<\/span>\n<ul>\n \t<li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\"><em>n<\/em> est le nombre total d&rsquo;exemples,<\/span><\/li>\n \t<li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\"><i>y<sub>i<\/sub><\/i> est la valeur r\u00e9elle pour l\u2019exemple i,<\/span><\/li>\n \t<li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\"><em>\u0177<sub>i<\/sub><\/em>\u200b est la valeur pr\u00e9dite par le mod\u00e8le pour cet exemple,<\/span><\/li>\n \t<li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\"><em>|y<sub>i<\/sub> &#8211; \u0177<sub>i<\/sub>|<\/em> est l\u2019erreur absolue pour cet exemple.<\/span><\/li>\n<\/ul>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-9410def elementor-widget elementor-widget-text-editor\" data-id=\"9410def\" 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;\">Pour reprendre notre exemple avec les prix de l\u2019immobilier, cette fonction de co\u00fbt nous permet d\u2019identifier la moyenne des \u00e9carts entre les prix pr\u00e9dits et les prix r\u00e9els. Et ce, en valeur absolue.\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-946d97a elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"946d97a\" 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\/formation-intelligence-artificielle\">\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\">Ma\u00eetriser la fonction de co\u00fbt en ML<\/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<div class=\"elementor-element elementor-element-730e8da elementor-widget elementor-widget-heading\" data-id=\"730e8da\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h4 class=\"elementor-heading-title elementor-size-default\"><font size=\"4\">Log Loss (ou logarithme de la perte)<\/font><\/h4>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-d449960 elementor-widget elementor-widget-text-editor\" data-id=\"d449960\" 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;\">Cette fonction de co\u00fbt est utilis\u00e9e pour <\/span><b>\u00e9valuer les mod\u00e8les de classification probabiliste<\/b><span style=\"font-weight: 400;\">, notamment dans les <a href=\"https:\/\/datascientest.com\/algorithme-de-classification-definition-et-principaux-modeles\">probl\u00e8mes de classification<\/a> binaire ou multiclasses. Concr\u00e8tement, elle mesure la \u00ab\u00a0distance\u00a0\u00bb entre les distributions des probabilit\u00e9s pr\u00e9dites et les vraies classes.<\/span><\/p><p><span style=\"font-weight: 400;\">Voici la formule math\u00e9matique : <\/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-3a338cf elementor-widget elementor-widget-image\" data-id=\"3a338cf\" 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=\"505\" height=\"71\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/12\/image3.png\" class=\"attachment-full size-full wp-image-259340\" alt=\"\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/12\/image3.png 505w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/12\/image3-300x42.png 300w\" sizes=\"(max-width: 505px) 100vw, 505px\" \/>\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-6acdfcf elementor-widget elementor-widget-text-editor\" data-id=\"6acdfcf\" 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;\">Et ses param\u00e8tres :\u00a0<\/span><\/p><ul><li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\"><em>n<\/em> est le nombre total d&rsquo;exemples dans le jeu de donn\u00e9es,<\/span><\/li><li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\"><em>y<sub>i<\/sub><\/em> est la classe r\u00e9elle de l\u2019exemple i (0 ou 1),<\/span><\/li><li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\"><em>\u0177<sub>i<\/sub><\/em> \u200best la probabilit\u00e9 pr\u00e9dite pour que <em>y<sub>i<\/sub>=1<\/em><\/span><\/li><li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\"><em>log<\/em> est le logarithme naturel.<\/span><\/li><\/ul>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-2b620ef elementor-widget elementor-widget-text-editor\" data-id=\"2b620ef\" 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;\">Si le mod\u00e8le pr\u00e9dit des probabilit\u00e9s proches des classes r\u00e9elles (par exemple, 0.9 pour une classe 1), la log loss est faible. En revanche, si le mod\u00e8le est tr\u00e8s \u00e9loign\u00e9 des probabilit\u00e9s correctes (0.1 pour une classe 1), la log loss est \u00e9lev\u00e9e. L\u00e0 encore, l\u2019objectif est d\u2019<\/span><b>avoir une fonction de co\u00fbt la plus faible possible<\/b><span style=\"font-weight: 400;\">, \u00e0 travers diff\u00e9rentes hypoth\u00e8ses et it\u00e9rations.\u00a0<\/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-5c8fcd2 e-flex e-con-boxed e-con\" data-id=\"5c8fcd2\" 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-30977fa elementor-widget elementor-widget-heading\" data-id=\"30977fa\" 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\">R\u00e9duisez les erreurs des mod\u00e8les d\u2019IA<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-f0af18b elementor-widget elementor-widget-text-editor\" data-id=\"f0af18b\" 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;\">Vous l\u2019aurez compris, la fonction co\u00fbt est incontournable pour concevoir des mod\u00e8les de machine learning performants. Mais elle est aussi complexe. Afin de bien comprendre ses diff\u00e9rents cas d\u2019application (mod\u00e8les de r\u00e9gression, de classification, r\u00e9seaux neuronaux, apprentissage par renforcement, \u2026), il est primordial de s\u2019y former. Et pour cela, il y a <strong>DataScientest<\/strong>. D\u00e9couvrez nos <a href=\"https:\/\/datascientest.com\/formation-data-scientist\">programmes en science de donn\u00e9es<\/a>.\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-92ba223 elementor-widget elementor-widget-image\" data-id=\"92ba223\" 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=\"1000\" height=\"571\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/12\/fonction-cout-machine-learning-datascientest-1.webp\" class=\"attachment-full size-full wp-image-259350\" alt=\"\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/12\/fonction-cout-machine-learning-datascientest-1.webp 1000w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/12\/fonction-cout-machine-learning-datascientest-1-300x171.webp 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/12\/fonction-cout-machine-learning-datascientest-1-768x439.webp 768w\" sizes=\"(max-width: 1000px) 100vw, 1000px\" \/>\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-725aacf elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"725aacf\" 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\/formation-intelligence-artificielle\">\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\">Se former avec DataScientest<\/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<div class=\"elementor-element elementor-element-98ae248 e-flex e-con-boxed e-con\" data-id=\"98ae248\" 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-c79c458 elementor-widget elementor-widget-html\" data-id=\"c79c458\" data-element_type=\"widget\" data-widget_type=\"html.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<script type=\"application\/ld+json\">\r\n{\r\n  \"@context\": \"https:\/\/schema.org\",\r\n  \"@type\": \"FAQPage\",\r\n  \"mainEntity\": [\r\n    {\r\n      \"@type\": \"Question\",\r\n      \"name\": \"Qu'est-ce qu'une fonction de co\u00fbt en apprentissage automatique ?\",\r\n      \"acceptedAnswer\": {\r\n        \"@type\": \"Answer\",\r\n        \"text\": \"La fonction de co\u00fbt est une mesure utilis\u00e9e en apprentissage automatique pour \u00e9valuer la diff\u00e9rence entre les pr\u00e9dictions d'un mod\u00e8le et les valeurs r\u00e9elles. Elle guide l'optimisation du mod\u00e8le en minimisant cette diff\u00e9rence.\"\r\n      }\r\n    },\r\n    {\r\n      \"@type\": \"Question\",\r\n      \"name\": \"Pourquoi utilise-t-on une fonction de co\u00fbt ?\",\r\n      \"acceptedAnswer\": {\r\n        \"@type\": \"Answer\",\r\n        \"text\": \"La fonction de co\u00fbt est utilis\u00e9e pour quantifier les erreurs d'un mod\u00e8le. En minimisant cette fonction, on am\u00e9liore la pr\u00e9cision des pr\u00e9dictions, ce qui permet d'obtenir un mod\u00e8le plus performant.\"\r\n      }\r\n    },\r\n    {\r\n      \"@type\": \"Question\",\r\n      \"name\": \"Quelles sont les fonctions de co\u00fbt courantes ?\",\r\n      \"acceptedAnswer\": {\r\n        \"@type\": \"Answer\",\r\n        \"text\": \"Les fonctions de co\u00fbt courantes incluent l'erreur quadratique moyenne (MSE), l'erreur absolue moyenne (MAE) pour les probl\u00e8mes de r\u00e9gression, et l'entropie crois\u00e9e pour les probl\u00e8mes de classification.\"\r\n      }\r\n    },\r\n    {\r\n      \"@type\": \"Question\",\r\n      \"name\": \"Comment choisir la bonne fonction de co\u00fbt ?\",\r\n      \"acceptedAnswer\": {\r\n        \"@type\": \"Answer\",\r\n        \"text\": \"Le choix de la fonction de co\u00fbt d\u00e9pend du type de probl\u00e8me : pour la r\u00e9gression, on privil\u00e9gie MSE ou MAE, tandis que pour la classification, on utilise g\u00e9n\u00e9ralement l'entropie crois\u00e9e ou des variantes adapt\u00e9es aux classes d\u00e9s\u00e9quilibr\u00e9es.\"\r\n      }\r\n    }\r\n  ]\r\n}\r\n<\/script>\r\n\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>Si l\u2019erreur est humaine, elle n\u2019est pas exclusive \u00e0 l\u2019Homme. Les algorithmes d\u2019apprentissage automatique peuvent aussi faire des erreurs. Mais \u00e0 la diff\u00e9rence de nous, pauvres humains, il est possible de minimiser. Notamment en utilisant la fonction de co\u00fbt qui permet d\u2019\u00e9valuer la performance d\u2019un mod\u00e8le de machine learning.&nbsp;Alors de quoi s\u2019agit-il exactement? Et comment [&hellip;]<\/p>\n","protected":false},"author":77,"featured_media":259336,"comment_status":"open","ping_status":"open","sticky":false,"template":"elementor_theme","format":"standard","meta":{"content-type":"","inline_featured_image":false},"categories":[2500],"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>La fonction de co\u00fbt en IA : Tout ce qu&#039;il faut savoir<\/title>\n<meta name=\"description\" content=\"Capable de mesurer l\u2019\u00e9cart entre des pr\u00e9dictions et des valeurs r\u00e9elles, la fonction de co\u00fbt est indispensable \u00e0 l&#039;entra\u00eenement des mod\u00e8les ML.\" \/>\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\/fonction-de-cout-tout-savoir\" \/>\n<meta property=\"og:locale\" content=\"fr_FR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"La fonction de co\u00fbt en IA : Tout ce qu&#039;il faut savoir\" \/>\n<meta property=\"og:url\" content=\"https:\/\/datascientest.com\/fonction-de-cout-tout-savoir\" \/>\n<meta property=\"og:site_name\" content=\"Formation Data Science | DataScientest.com\" \/>\n<meta property=\"article:published_time\" content=\"2025-01-14T06:30:00+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2025-01-10T13:01:09+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/12\/fonction-cout-machine-learning-datascientest.webp\" \/>\n\t<meta property=\"og:image:width\" content=\"1200\" \/>\n\t<meta property=\"og:image:height\" content=\"686\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/webp\" \/>\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\/fonction-de-cout-tout-savoir\",\"url\":\"https:\/\/datascientest.com\/fonction-de-cout-tout-savoir\",\"name\":\"La fonction de co\u00fbt en IA : Tout ce qu'il faut savoir\",\"isPartOf\":{\"@id\":\"https:\/\/datascientest.com\/#website\"},\"datePublished\":\"2025-01-14T06:30:00+00:00\",\"dateModified\":\"2025-01-10T13:01:09+00:00\",\"author\":{\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/6eca16f63daaafcbc96e15549b9d4ee1\"},\"description\":\"Capable de mesurer l\u2019\u00e9cart entre des pr\u00e9dictions et des valeurs r\u00e9elles, la fonction de co\u00fbt est indispensable \u00e0 l'entra\u00eenement des mod\u00e8les ML.\",\"inLanguage\":\"fr-FR\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/datascientest.com\/fonction-de-cout-tout-savoir\"]}]},{\"@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":"La fonction de co\u00fbt en IA : Tout ce qu'il faut savoir","description":"Capable de mesurer l\u2019\u00e9cart entre des pr\u00e9dictions et des valeurs r\u00e9elles, la fonction de co\u00fbt est indispensable \u00e0 l'entra\u00eenement des mod\u00e8les ML.","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\/fonction-de-cout-tout-savoir","og_locale":"fr_FR","og_type":"article","og_title":"La fonction de co\u00fbt en IA : Tout ce qu'il faut savoir","og_url":"https:\/\/datascientest.com\/fonction-de-cout-tout-savoir","og_site_name":"Formation Data Science | DataScientest.com","article_published_time":"2025-01-14T06:30:00+00:00","article_modified_time":"2025-01-10T13:01:09+00:00","og_image":[{"width":1200,"height":686,"url":"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/12\/fonction-cout-machine-learning-datascientest.webp","type":"image\/webp"}],"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\/fonction-de-cout-tout-savoir","url":"https:\/\/datascientest.com\/fonction-de-cout-tout-savoir","name":"La fonction de co\u00fbt en IA : Tout ce qu'il faut savoir","isPartOf":{"@id":"https:\/\/datascientest.com\/#website"},"datePublished":"2025-01-14T06:30:00+00:00","dateModified":"2025-01-10T13:01:09+00:00","author":{"@id":"https:\/\/datascientest.com\/#\/schema\/person\/6eca16f63daaafcbc96e15549b9d4ee1"},"description":"Capable de mesurer l\u2019\u00e9cart entre des pr\u00e9dictions et des valeurs r\u00e9elles, la fonction de co\u00fbt est indispensable \u00e0 l'entra\u00eenement des mod\u00e8les ML.","inLanguage":"fr-FR","potentialAction":[{"@type":"ReadAction","target":["https:\/\/datascientest.com\/fonction-de-cout-tout-savoir"]}]},{"@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\/259331"}],"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=259331"}],"version-history":[{"count":9,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts\/259331\/revisions"}],"predecessor-version":[{"id":261015,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts\/259331\/revisions\/261015"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/media\/259336"}],"wp:attachment":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/media?parent=259331"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/categories?post=259331"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/tags?post=259331"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}