{"id":112611,"date":"2021-11-23T09:56:30","date_gmt":"2021-11-23T08:56:30","guid":{"rendered":"https:\/\/datascientest.com\/?p=112611"},"modified":"2023-11-09T14:05:42","modified_gmt":"2023-11-09T14:05:42","slug":"adversarial-training-tout-savoir","status":"publish","type":"post","link":"https:\/\/datascientest.com\/adversarial-training-tout-savoir","title":{"rendered":"Adversarial Training : qu&rsquo;est-ce que c&rsquo;est ?"},"content":{"rendered":"\t\t<div data-elementor-type=\"wp-post\" data-elementor-id=\"112611\" class=\"elementor elementor-112611\" data-elementor-post-type=\"post\">\n\t\t\t\t\t\t\t\t\t<section class=\"elementor-section elementor-top-section elementor-element elementor-element-0057b97 elementor-section-boxed elementor-section-height-default elementor-section-height-default\" data-id=\"0057b97\" data-element_type=\"section\">\n\t\t\t\t\t\t<div class=\"elementor-container elementor-column-gap-default\">\n\t\t\t\t\t<div class=\"elementor-column elementor-col-100 elementor-top-column elementor-element elementor-element-e4a5090\" data-id=\"e4a5090\" data-element_type=\"column\">\n\t\t\t<div class=\"elementor-widget-wrap elementor-element-populated\">\n\t\t\t\t\t\t\t\t<div class=\"elementor-element elementor-element-9f6a110 elementor-widget elementor-widget-heading\" data-id=\"9f6a110\" 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\">Depuis les ann\u00e9es 2010, gr\u00e2ce aux progr\u00e8s du Machine Learning et notamment du Deep Learning avec les r\u00e9seaux neuronaux profonds, les erreurs sont devenues de plus en plus rares. Aujourd\u2019hui elles sont m\u00eame tr\u00e8s exceptionnelles. Cependant, ces mod\u00e8les continuent parfois de se tromper, sans que les chercheurs r\u00e9ussissent \u00e0 d\u00e9velopper des syst\u00e8mes de d\u00e9fense efficaces. <\/h2>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-5e0f0ef elementor-widget elementor-widget-text-editor\" data-id=\"5e0f0ef\" 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=\"color: #000000;\">Les <a href=\"https:\/\/datascientest.com\/adversarial-examples\">Adver<\/a><a href=\"https:\/\/datascientest.com\/adversarial-examples\">sarial<\/a><a href=\"https:\/\/datascientest.com\/adversarial-examples\"> Examples<\/a> ou exemples contradictoires font partie de ces inputs que le mod\u00e8le va mal classifier. Face \u00e0 cela, une technique de d\u00e9fense, nomm\u00e9e Adversarial Training ou entra\u00eenement contradictoire, a \u00e9t\u00e9 d\u00e9velopp\u00e9e. <\/span><span style=\"color: #000000;\">Mais comment fonctionne cette technique de d\u00e9fense ? Est-elle vraiment efficace ? <\/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-60e7812 elementor-widget elementor-widget-heading\" data-id=\"60e7812\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h2 class=\"elementor-heading-title elementor-size-default\">Qu\u2019est-ce qu\u2019un Adversarial Example ?<\/h2>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-cd332c3 elementor-widget elementor-widget-text-editor\" data-id=\"cd332c3\" 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; color: #000000;\">Le Adversarial Training est une technique qui a \u00e9t\u00e9 d\u00e9velopp\u00e9e pour prot\u00e9ger les mod\u00e8les de Machine Learning face aux Adversarial Examples. Rappelons bri\u00e8vement en quoi consistent les Adversarial Examples. Ce sont des inputs tr\u00e8s l\u00e9g\u00e8rement et judicieusement perturb\u00e9s (comme une image, un texte, un son), d\u2019une mani\u00e8re imperceptible pour l\u2019humain, mais qui vont mal \u00eatre classifi\u00e9s par un mod\u00e8le de machine learning. <\/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-314e478 elementor-widget elementor-widget-image\" data-id=\"314e478\" 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<figure class=\"wp-caption\">\n\t\t\t\t\t\t\t\t\t\t\t<a href=\"https:\/\/openai.com\/blog\/adversarial-example-research\/\">\n\t\t\t\t\t\t\t<img decoding=\"async\" width=\"512\" height=\"203\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2021\/10\/unnamed-1.png\" class=\"attachment-large size-large wp-image-112618\" alt=\"Classification d&#039;une image de panda gr\u00e2ce \u00e0 un adversarial network\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2021\/10\/unnamed-1.png 512w, https:\/\/datascientest.com\/wp-content\/uploads\/2021\/10\/unnamed-1-300x119.png 300w\" sizes=\"(max-width: 512px) 100vw, 512px\" \/>\t\t\t\t\t\t\t\t<\/a>\n\t\t\t\t\t\t\t\t\t\t\t<figcaption class=\"widget-image-caption wp-caption-text\">Classification d'une image de panda gr\u00e2ce \u00e0 un adversarial network<\/figcaption>\n\t\t\t\t\t\t\t\t\t\t<\/figure>\n\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-1e37af5 elementor-widget elementor-widget-text-editor\" data-id=\"1e37af5\" 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; color: #000000;\">Ce qui est stup\u00e9fiant avec ces attaques, c\u2019est l\u2019assurance qu\u2019a le mod\u00e8le dans sa fausse pr\u00e9diction. L\u2019exemple ci-dessus le montre bien : alors que le mod\u00e8le admet un taux de confiance de seulement 57,7% pour la pr\u00e9diction qui est juste, il va admettre un taux de confiance tr\u00e8s \u00e9lev\u00e9 de 99,3% pour la pr\u00e9diction fausse. Si vous voulez en savoir plus sur ces attaques \u00e9tonnantes allez voir l\u2019article d\u00e9di\u00e9 : <a href=\"https:\/\/datascientest.com\/adversarial-attack-quest-ce-que-cest-et-comment-proteger-lia-contre-cette-menace\">https:\/\/datascientest.com\/adversarial-attack-quest-ce-que-cest-et-comment-proteger-lia-contre-cette-menace<\/a><\/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-34c03d0 elementor-widget elementor-widget-text-editor\" data-id=\"34c03d0\" 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=\"color: #000000;\">Ces attaques sont tr\u00e8s probl\u00e9matiques. Par exemple, <a href=\"https:\/\/www.science.org\/doi\/10.1126\/science.aaw4399\">un article publi\u00e9 dans Science en 2019<\/a> par des chercheurs de Harvard et du MIT montre comment les syst\u00e8mes d&rsquo;IA m\u00e9dicaux pourraient \u00eatre vuln\u00e9rables aux attaques adverses. C\u2019est pourquoi il est n\u00e9cessaire de se d\u00e9fendre. C\u2019est l\u00e0 qu\u2019intervient le Adversarial Training. Il s\u2019agit, avec la \u00ab\u00a0Defensive Distillation\u00a0\u00bb, de la principale technique pour se prot\u00e9ger de ces attaques. <\/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<\/div>\n\t\t\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t<section class=\"elementor-section elementor-top-section elementor-element elementor-element-1c71546 elementor-section-boxed elementor-section-height-default elementor-section-height-default\" data-id=\"1c71546\" data-element_type=\"section\">\n\t\t\t\t\t\t<div class=\"elementor-container elementor-column-gap-default\">\n\t\t\t\t\t<div class=\"elementor-column elementor-col-100 elementor-top-column elementor-element elementor-element-9eaefe3\" data-id=\"9eaefe3\" data-element_type=\"column\">\n\t\t\t<div class=\"elementor-widget-wrap elementor-element-populated\">\n\t\t\t\t\t\t\t\t<div class=\"elementor-element elementor-element-c0771fd elementor-widget elementor-widget-heading\" data-id=\"c0771fd\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h2 class=\"elementor-heading-title elementor-size-default\">Comment fonctionne l'Adversarial Training ?<\/h2>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-d1c4138 elementor-widget elementor-widget-text-editor\" data-id=\"d1c4138\" 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; color: #000000;\">Comment cette technique fonctionne-t-elle\u00a0? Il s\u2019agit en fait de r\u00e9entra\u00eener le mod\u00e8le de Machine Learning avec de nombreux Adversarial Examples. En effet, dans la phase d\u2019entra\u00eenement d\u2019un mod\u00e8le pr\u00e9dictif, si l\u2019input est mal classifi\u00e9 par le mod\u00e8le de Machine Learning, l\u2019algorithme apprend de ses erreurs et r\u00e9ajuste ses param\u00e8tres dans le but de ne plus les commettre.<\/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-ac86d38 elementor-widget elementor-widget-image\" data-id=\"ac86d38\" 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<figure class=\"wp-caption\">\n\t\t\t\t\t\t\t\t\t\t\t<a href=\"https:\/\/www.futura-sciences.com\/tech\/definitions\/intelligence-artificielle-deep-learning-17262\/\">\n\t\t\t\t\t\t\t<img decoding=\"async\" width=\"512\" height=\"275\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2021\/10\/unnamed.jpeg\" class=\"attachment-large size-large wp-image-112659\" alt=\"illustration du fonctionnement de l&#039;adversarial training\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2021\/10\/unnamed.jpeg 512w, https:\/\/datascientest.com\/wp-content\/uploads\/2021\/10\/unnamed-300x161.jpeg 300w\" sizes=\"(max-width: 512px) 100vw, 512px\" \/>\t\t\t\t\t\t\t\t<\/a>\n\t\t\t\t\t\t\t\t\t\t\t<figcaption class=\"widget-image-caption wp-caption-text\">illustration du fonctionnement de l'adversarial training<\/figcaption>\n\t\t\t\t\t\t\t\t\t\t<\/figure>\n\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-248a78f elementor-widget elementor-widget-text-editor\" data-id=\"248a78f\" 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; color: #000000;\">Ainsi, apr\u00e8s avoir entra\u00een\u00e9 une premi\u00e8re fois le mod\u00e8le, les concepteurs du mod\u00e8le vont g\u00e9n\u00e9rer de nombreux Adversarial Examples. Ils vont confronter leur propre mod\u00e8le \u00e0 ces exemples contradictoires pour qu\u2019il ne commette plus ces erreurs.<\/span><\/p><p><span style=\"font-weight: 400; color: #000000;\">Si cette m\u00e9thode va d\u00e9fendre les mod\u00e8les de Machine Learning contre certains Adversarial Examples, permet-elle de g\u00e9n\u00e9raliser la robustesse du mod\u00e8le \u00e0 tous les Adversarial Examples\u00a0? La r\u00e9ponse est non. Cette approche est globalement insuffisante pour arr\u00eater toutes les attaques, car l&rsquo;\u00e9ventail des attaques possibles est trop large et ne peut \u00eatre g\u00e9n\u00e9r\u00e9 \u00e0 l&rsquo;avance. Ainsi, il s\u2019agit souvent d\u2019une course entre les hackers g\u00e9n\u00e8rant de nouveaux adversarial examples, et les concepteurs s\u2019en prot\u00e8geant le plus vite possible.\u00a0<\/span><\/p><p><span style=\"font-weight: 400; color: #000000;\">De mani\u00e8re plus g\u00e9n\u00e9rale, il est tr\u00e8s difficile de prot\u00e9ger les mod\u00e8les contre les adversarial examples, parce qu\u2019il est quasiment impossible de construire un mod\u00e8le th\u00e9orique de l\u2019\u00e9laboration des ces exemples. Il s\u2019agirait de r\u00e9soudre des probl\u00e8mes d\u2019optimisation particuli\u00e8rement complexes, et nous ne disposons pas des outils th\u00e9oriques n\u00e9cessaires.\u00a0<\/span><\/p><p><span style=\"font-weight: 400; color: #000000;\">Toutes les strat\u00e9gies test\u00e9es jusqu&rsquo;\u00e0 pr\u00e9sent \u00e9chouent parce qu&rsquo;elles ne sont pas adaptatives : elles peuvent bloquer un type d&rsquo;attaque, mais laissent une autre vuln\u00e9rabilit\u00e9 ouverte \u00e0 un attaquant qui conna\u00eet la d\u00e9fense utilis\u00e9e. La conception d&rsquo;une d\u00e9fense capable de prot\u00e9ger contre un hacker puissant et adaptatif est un domaine de recherche important.<\/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-a59ccbf elementor-widget elementor-widget-text-editor\" data-id=\"a59ccbf\" 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; color: #000000;\">Pour conclure, l\u2019Adversarial Training \u00e9choue globalement \u00e0 prot\u00e9ger les mod\u00e8les de Machine Learning contre les <\/span><span style=\"color: #000000;\"><b>Adversarial Attacks<\/b><\/span><span style=\"font-weight: 400; color: #000000;\">. S\u2019il fallait retenir une raison, c\u2019est parce que cette technique propose une d\u00e9fense contre une s\u00e9rie d\u2019attaques particuli\u00e8res, sans parvenir \u00e0 d\u00e9gager une m\u00e9thode g\u00e9n\u00e9ralis\u00e9e.\u00a0<\/span><\/p><p><span style=\"color: #000000;\">Vous voulez en savoir plus sur les d\u00e9fis de l\u2019intelligence artificielle\u00a0? Envie de ma\u00eetriser les techniques de Deep Learning \u00e9voqu\u00e9es dans cet article ? Renseignez-vous sur <a href=\"https:\/\/datascientest.com\/formation-machine-learning-engineer\">notre formation de Machine Learning Engineer<\/a>.<\/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-b54d019 elementor-align-center elementor-widget elementor-widget-button\" data-id=\"b54d019\" 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-machine-learning-engineer\">\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 le cursus Machine Learning Engineer<\/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<\/div>\n\t\t\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t\t\t\t<\/div>\n\t\t","protected":false},"excerpt":{"rendered":"<p>Depuis les ann\u00e9es 2010, gr\u00e2ce aux progr\u00e8s du Machine Learning et notamment du Deep Learning avec les r\u00e9seaux neuronaux profonds, les erreurs sont devenues de plus en plus rares. Aujourd\u2019hui elles sont m\u00eame tr\u00e8s exceptionnelles. Cependant, ces mod\u00e8les continuent parfois de se tromper, sans que les chercheurs r\u00e9ussissent \u00e0 d\u00e9velopper des syst\u00e8mes de d\u00e9fense efficaces. [&hellip;]<\/p>\n","protected":false},"author":78,"featured_media":127402,"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>Adversarial Training : qu&#039;est-ce que c&#039;est ?<\/title>\n<meta name=\"description\" content=\"Depuis les ann\u00e9es 2010, gr\u00e2ce aux progr\u00e8s du Machine Learning et notamment du Deep Learning avec les r\u00e9seaux neuronaux profonds, les erreurs sont devenues de plus en plus rares.\" \/>\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\/adversarial-training-tout-savoir\" \/>\n<meta property=\"og:locale\" content=\"fr_FR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Adversarial Training : qu&#039;est-ce que c&#039;est ?\" \/>\n<meta property=\"og:url\" content=\"https:\/\/datascientest.com\/adversarial-training-tout-savoir\" \/>\n<meta property=\"og:site_name\" content=\"Formation Data Science | DataScientest.com\" \/>\n<meta property=\"article:published_time\" content=\"2021-11-23T08:56:30+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2023-11-09T14:05:42+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/datascientest.com\/wp-content\/uploads\/2021\/11\/illu_adversarial_training-71.png\" \/>\n\t<meta property=\"og:image:width\" content=\"2560\" \/>\n\t<meta property=\"og:image:height\" content=\"1405\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"J\u00e9r\u00e9my Robert\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u00c9crit par\" \/>\n\t<meta name=\"twitter:data1\" content=\"J\u00e9r\u00e9my Robert\" \/>\n\t<meta name=\"twitter:label2\" content=\"Dur\u00e9e de lecture estim\u00e9e\" \/>\n\t<meta name=\"twitter:data2\" content=\"4 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/datascientest.com\/adversarial-training-tout-savoir\",\"url\":\"https:\/\/datascientest.com\/adversarial-training-tout-savoir\",\"name\":\"Adversarial Training : qu'est-ce que c'est ?\",\"isPartOf\":{\"@id\":\"https:\/\/datascientest.com\/#website\"},\"datePublished\":\"2021-11-23T08:56:30+00:00\",\"dateModified\":\"2023-11-09T14:05:42+00:00\",\"author\":{\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/7af2361c48e24e79a8aa2360e6e74925\"},\"description\":\"Depuis les ann\u00e9es 2010, gr\u00e2ce aux progr\u00e8s du Machine Learning et notamment du Deep Learning avec les r\u00e9seaux neuronaux profonds, les erreurs sont devenues de plus en plus rares.\",\"inLanguage\":\"fr-FR\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/datascientest.com\/adversarial-training-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\/7af2361c48e24e79a8aa2360e6e74925\",\"name\":\"J\u00e9r\u00e9my Robert\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"fr-FR\",\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/jeremyrobert.jpeg\",\"contentUrl\":\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/jeremyrobert.jpeg\",\"caption\":\"J\u00e9r\u00e9my Robert\"},\"description\":\"J\u00e9r\u00e9my Robert est actuellement Directeur p\u00e9dagogique B2B chez DataScientest.com. Avant de rejoindre DataScientest, il a travaill\u00e9 en tant que Consultant Business Intelligence chez DEAPDATA, o\u00f9 il a g\u00e9r\u00e9 des projets dans les domaines de l'assurance, bancaire et industrie. Il a \u00e9galement \u00e9t\u00e9 Co-fondateur et pr\u00e9sident de Tripzer. En tant que consultant ind\u00e9pendant, J\u00e9r\u00e9my a propos\u00e9 des formations sur MicroStrategy pour divers niveaux professionnels. Il est dipl\u00f4m\u00e9 de l'Universit\u00e9 Paris Cit\u00e9 avec un Master MIAGE, combinant comp\u00e9tences en informatique et gestion.\",\"sameAs\":[\"https:\/\/www.linkedin.com\/in\/jeremyrobert1\/\",\"Data Science\"],\"url\":\"https:\/\/datascientest.com\/author\/jeremyrobert\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"Adversarial Training : qu'est-ce que c'est ?","description":"Depuis les ann\u00e9es 2010, gr\u00e2ce aux progr\u00e8s du Machine Learning et notamment du Deep Learning avec les r\u00e9seaux neuronaux profonds, les erreurs sont devenues de plus en plus rares.","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\/adversarial-training-tout-savoir","og_locale":"fr_FR","og_type":"article","og_title":"Adversarial Training : qu'est-ce que c'est ?","og_url":"https:\/\/datascientest.com\/adversarial-training-tout-savoir","og_site_name":"Formation Data Science | DataScientest.com","article_published_time":"2021-11-23T08:56:30+00:00","article_modified_time":"2023-11-09T14:05:42+00:00","og_image":[{"width":2560,"height":1405,"url":"https:\/\/datascientest.com\/wp-content\/uploads\/2021\/11\/illu_adversarial_training-71.png","type":"image\/png"}],"author":"J\u00e9r\u00e9my Robert","twitter_card":"summary_large_image","twitter_misc":{"\u00c9crit par":"J\u00e9r\u00e9my Robert","Dur\u00e9e de lecture estim\u00e9e":"4 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/datascientest.com\/adversarial-training-tout-savoir","url":"https:\/\/datascientest.com\/adversarial-training-tout-savoir","name":"Adversarial Training : qu'est-ce que c'est ?","isPartOf":{"@id":"https:\/\/datascientest.com\/#website"},"datePublished":"2021-11-23T08:56:30+00:00","dateModified":"2023-11-09T14:05:42+00:00","author":{"@id":"https:\/\/datascientest.com\/#\/schema\/person\/7af2361c48e24e79a8aa2360e6e74925"},"description":"Depuis les ann\u00e9es 2010, gr\u00e2ce aux progr\u00e8s du Machine Learning et notamment du Deep Learning avec les r\u00e9seaux neuronaux profonds, les erreurs sont devenues de plus en plus rares.","inLanguage":"fr-FR","potentialAction":[{"@type":"ReadAction","target":["https:\/\/datascientest.com\/adversarial-training-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\/7af2361c48e24e79a8aa2360e6e74925","name":"J\u00e9r\u00e9my Robert","image":{"@type":"ImageObject","inLanguage":"fr-FR","@id":"https:\/\/datascientest.com\/#\/schema\/person\/image\/","url":"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/jeremyrobert.jpeg","contentUrl":"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/jeremyrobert.jpeg","caption":"J\u00e9r\u00e9my Robert"},"description":"J\u00e9r\u00e9my Robert est actuellement Directeur p\u00e9dagogique B2B chez DataScientest.com. Avant de rejoindre DataScientest, il a travaill\u00e9 en tant que Consultant Business Intelligence chez DEAPDATA, o\u00f9 il a g\u00e9r\u00e9 des projets dans les domaines de l'assurance, bancaire et industrie. Il a \u00e9galement \u00e9t\u00e9 Co-fondateur et pr\u00e9sident de Tripzer. En tant que consultant ind\u00e9pendant, J\u00e9r\u00e9my a propos\u00e9 des formations sur MicroStrategy pour divers niveaux professionnels. Il est dipl\u00f4m\u00e9 de l'Universit\u00e9 Paris Cit\u00e9 avec un Master MIAGE, combinant comp\u00e9tences en informatique et gestion.","sameAs":["https:\/\/www.linkedin.com\/in\/jeremyrobert1\/","Data Science"],"url":"https:\/\/datascientest.com\/author\/jeremyrobert"}]}},"_links":{"self":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts\/112611"}],"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\/78"}],"replies":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/comments?post=112611"}],"version-history":[{"count":0,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts\/112611\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/media\/127402"}],"wp:attachment":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/media?parent=112611"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/categories?post=112611"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/tags?post=112611"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}