{"id":7352,"date":"2026-07-09T10:10:04","date_gmt":"2026-07-09T08:10:04","guid":{"rendered":"https:\/\/francestat.com\/?page_id=7352"},"modified":"2026-07-09T10:54:56","modified_gmt":"2026-07-09T08:54:56","slug":"uniwin-gboost","status":"publish","type":"page","link":"https:\/\/francestat.com\/index.php\/uniwin-gboost\/","title":{"rendered":"Uniwin – GBOOST"},"content":{"rendered":"

[vc_row][vc_column]

<\/div>

UNIWIN - Renforcement du gradient<\/h2><\/div><\/div>[\/vc_column][\/vc_row][vc_row][vc_column][vc_empty_space][\/vc_column][\/vc_row][vc_row][vc_column][vc_column_text css=\u00a0\u00bb\u00a0\u00bb]La proc\u00e9dure Renforcement du gradient (gradient boosting) impl\u00e9mente une m\u00e9thode d’apprentissage machine pour pr\u00e9voir des observations \u00e0 partir de donn\u00e9es. Elle cr\u00e9e des mod\u00e8les de deux formes :<\/p>\n
    \n
  1. Mod\u00e8les de classement qui divisent des observations en classes en se basant sur les caract\u00e9ristiques observ\u00e9es.<\/li>\n
  2. Mod\u00e8les de r\u00e9gression qui pr\u00e9voient la valeur d’une variable cible.<\/li>\n<\/ol>\n

    Elle consiste \u00e0 combiner progressivement des mod\u00e8les faibles bas\u00e9s sur des arbres de d\u00e9cision ou de r\u00e9gression pour former un mod\u00e8le pr\u00e9dictif performant en corrigeant les erreurs r\u00e9siduelles de mani\u00e8re it\u00e9rative. Cette m\u00e9thode repose sur une optimisation s\u00e9quentielle o\u00f9 chaque nouveau mod\u00e8le ajuste ses param\u00e8tres en minimisant une fonction de perte via le calcul des gradients.<\/p>\n

    Les observations sont classiquement divis\u00e9es en trois jeux : un jeu d’apprentissage utilis\u00e9 pour construire le mod\u00e8le, un jeu de validation, pour lequel la classe ou la valeur est connue, utilis\u00e9 pour valider le mod\u00e8le et un jeu de pr\u00e9vision, pour lequel la classe ou la valeur n’est pas connue, utilis\u00e9 pour faire les pr\u00e9visions d\u00e9sir\u00e9es.<\/p>\n

    La variable cible et les caract\u00e9ristiques pr\u00e9dictives peuvent \u00eatre qualitatives ou quantitatives.<\/p>\n

    Cette proc\u00e9dure est bas\u00e9e sur le package R \u2018gbm\u2019 (Gradient Boosting Machine).[\/vc_column_text][\/vc_column][\/vc_row][vc_row][vc_column][vc_single_image image=\u00a0\u00bb7371″ img_size=\u00a0\u00bblarge\u00a0\u00bb alignment=\u00a0\u00bbcenter\u00a0\u00bb css=\u00a0\u00bb\u00a0\u00bb][vc_empty_space height=\u00a0\u00bb5px\u00a0\u00bb][vc_column_text css=\u00a0\u00bb\u00a0\u00bb]Tableaux<\/strong><\/p>\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
    Erreurs (apprentissage, validation crois\u00e9e) et am\u00e9liorations OOB<\/span><\/td>\n<\/tr>\n
    Importance des variables explicatives<\/span><\/td>\n<\/tr>\n
    Valeurs observ\u00e9es, pr\u00e9vues et probabilit\u00e9s (classement, apprentissage)<\/span><\/td>\n<\/tr>\n
    Matrice de confusion (classement, apprentissage)<\/span><\/td>\n<\/tr>\n
    Sensibilit\u00e9, sp\u00e9cificit\u00e9 (classement, apprentissage)<\/span><\/td>\n<\/tr>\n
    Valeurs observ\u00e9es, pr\u00e9vues et probabilit\u00e9s (classement, validation)<\/span><\/td>\n<\/tr>\n
    Matrice de confusion (classement, validation)<\/td>\n<\/tr>\n
    Sensibilit\u00e9, sp\u00e9cificit\u00e9 (classement, validation)<\/td>\n<\/tr>\n
    Valeurs pr\u00e9vues et probabilit\u00e9s (classement, pr\u00e9vision)<\/span><\/td>\n<\/tr>\n
    Valeurs observ\u00e9es, pr\u00e9vues et r\u00e9sidus (r\u00e9gression, apprentissage)<\/td>\n<\/tr>\n
    Valeurs observ\u00e9es, pr\u00e9vues et r\u00e9sidus (r\u00e9gression, validation)<\/span><\/td>\n<\/tr>\n
    Valeurs pr\u00e9vues (r\u00e9gression, pr\u00e9vision)<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

    Graphiques<\/strong><\/p>\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
    Graphique des erreurs (apprentissage, validation crois\u00e9e)<\/span><\/td>\n<\/tr>\n
    Graphique de l’importance des variables explicatives<\/span><\/td>\n<\/tr>\n
    Nuage de points (apprentissage)<\/td>\n<\/tr>\n
    Nuage de points (validation)<\/td>\n<\/tr>\n
    Nuage de points (pr\u00e9vision)<\/td>\n<\/tr>\n
    Graphique des fronti\u00e8res (apprentissage)<\/td>\n<\/tr>\n
    Graphique des fronti\u00e8res (validation)<\/span><\/td>\n<\/tr>\n
    Diagramme de la matrice de confusion (classement, apprentissage)<\/span><\/td>\n<\/tr>\n
    Diagramme de la matrice de confusion (classement, validation)<\/span><\/td>\n<\/tr>\n
    Courbes ROC (classement, apprentissage)\u00a0<\/span><\/td>\n<\/tr>\n
    Courbes ROC (classement, validation)<\/span><\/td>\n<\/tr>\n
    Pr\u00e9vus versus observ\u00e9s (r\u00e9gression, apprentissage)<\/span><\/td>\n<\/tr>\n
    R\u00e9sidus versus pr\u00e9vus (r\u00e9gression, apprentissage)<\/span><\/td>\n<\/tr>\n
    R\u00e9sidus versus pr\u00e9vus (r\u00e9gression, validation)<\/span><\/td>\n<\/tr>\n
    Pr\u00e9vus versus observ\u00e9s (r\u00e9gression, validation)<\/span><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

    [\/vc_column_text][vc_empty_space height=\u00a0\u00bb5px\u00a0\u00bb][\/vc_column][\/vc_row][vc_row][vc_column][vc_btn title=\u00a0\u00bbConsulter la documentation compl\u00e8te\u00a0\u00bb align=\u00a0\u00bbcenter\u00a0\u00bb css=\u00a0\u00bb\u00a0\u00bb link=\u00a0\u00bburl:http%3A%2F%2Fwww.francestat.com%2Ftelecharg%2FUniwin%2Fpdf%2FRenforcement%20du%20gradient.pdf|title:UNIWIN%20-%20GBOOST\u00a0\u00bb][\/vc_column][\/vc_row]<\/p>\n<\/div>","protected":false},"excerpt":{"rendered":"

    [vc_row][vc_column][\/vc_column][\/vc_row][vc_row][vc_column][vc_empty_space][\/vc_column][\/vc_row][vc_row][vc_column][vc_column_text css=\u00a0\u00bb\u00a0\u00bb]La proc\u00e9dure Renforcement du gradient (gradient boosting) impl\u00e9mente une m\u00e9thode d’apprentissage machine pour pr\u00e9voir des observations \u00e0 partir de donn\u00e9es. Elle cr\u00e9e des mod\u00e8les de deux formes : Mod\u00e8les de classement qui divisent des observations en classes en se basant sur les caract\u00e9ristiques observ\u00e9es. Mod\u00e8les de r\u00e9gression qui pr\u00e9voient la valeur d’une variable cible.…<\/p>\n","protected":false},"author":1,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"class_list":["post-7352","page","type-page","status-publish","hentry","description-off"],"yoast_head":"\nUniwin - GBOOST - FRANCESTAT<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/francestat.com\/index.php\/uniwin-gboost\/\" \/>\n<meta property=\"og:locale\" content=\"fr_FR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Uniwin - GBOOST - FRANCESTAT\" \/>\n<meta property=\"og:description\" content=\"[vc_row][vc_column][\/vc_column][\/vc_row][vc_row][vc_column][vc_empty_space][\/vc_column][\/vc_row][vc_row][vc_column][vc_column_text css=\u00a0\u00bb\u00a0\u00bb]La proc\u00e9dure Renforcement du gradient (gradient boosting) impl\u00e9mente une m\u00e9thode d’apprentissage machine pour pr\u00e9voir des observations \u00e0 partir de donn\u00e9es. Elle cr\u00e9e des mod\u00e8les de deux formes : Mod\u00e8les de classement qui divisent des observations en classes en se basant sur les caract\u00e9ristiques observ\u00e9es. Mod\u00e8les de r\u00e9gression qui pr\u00e9voient la valeur d’une variable cible.…\" \/>\n<meta property=\"og:url\" content=\"https:\/\/francestat.com\/index.php\/uniwin-gboost\/\" \/>\n<meta property=\"og:site_name\" content=\"FRANCESTAT\" \/>\n<meta property=\"article:modified_time\" content=\"2026-07-09T08:54:56+00:00\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Dur\u00e9e de lecture estim\u00e9e\" \/>\n\t<meta name=\"twitter:data1\" content=\"2 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/francestat.com\\\/index.php\\\/uniwin-gboost\\\/\",\"url\":\"https:\\\/\\\/francestat.com\\\/index.php\\\/uniwin-gboost\\\/\",\"name\":\"Uniwin - GBOOST - FRANCESTAT\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/francestat.com\\\/#website\"},\"datePublished\":\"2026-07-09T08:10:04+00:00\",\"dateModified\":\"2026-07-09T08:54:56+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/francestat.com\\\/index.php\\\/uniwin-gboost\\\/#breadcrumb\"},\"inLanguage\":\"fr-FR\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/francestat.com\\\/index.php\\\/uniwin-gboost\\\/\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/francestat.com\\\/index.php\\\/uniwin-gboost\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Accueil\",\"item\":\"https:\\\/\\\/francestat.com\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Uniwin – GBOOST\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/francestat.com\\\/#website\",\"url\":\"https:\\\/\\\/francestat.com\\\/\",\"name\":\"FRANCESTAT\",\"description\":\"Logiciels, formations et services statistiques\",\"publisher\":{\"@id\":\"https:\\\/\\\/francestat.com\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/francestat.com\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"fr-FR\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/francestat.com\\\/#organization\",\"name\":\"FRANCESTAT\",\"url\":\"https:\\\/\\\/francestat.com\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"fr-FR\",\"@id\":\"https:\\\/\\\/francestat.com\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/francestat.com\\\/wp-content\\\/uploads\\\/2018\\\/05\\\/logo_horizontal_small_francestat.jpg\",\"contentUrl\":\"https:\\\/\\\/francestat.com\\\/wp-content\\\/uploads\\\/2018\\\/05\\\/logo_horizontal_small_francestat.jpg\",\"width\":155,\"height\":51,\"caption\":\"FRANCESTAT\"},\"image\":{\"@id\":\"https:\\\/\\\/francestat.com\\\/#\\\/schema\\\/logo\\\/image\\\/\"}}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Uniwin - GBOOST - FRANCESTAT","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:\/\/francestat.com\/index.php\/uniwin-gboost\/","og_locale":"fr_FR","og_type":"article","og_title":"Uniwin - GBOOST - FRANCESTAT","og_description":"[vc_row][vc_column][\/vc_column][\/vc_row][vc_row][vc_column][vc_empty_space][\/vc_column][\/vc_row][vc_row][vc_column][vc_column_text css=\u00a0\u00bb\u00a0\u00bb]La proc\u00e9dure Renforcement du gradient (gradient boosting) impl\u00e9mente une m\u00e9thode d’apprentissage machine pour pr\u00e9voir des observations \u00e0 partir de donn\u00e9es. Elle cr\u00e9e des mod\u00e8les de deux formes : Mod\u00e8les de classement qui divisent des observations en classes en se basant sur les caract\u00e9ristiques observ\u00e9es. Mod\u00e8les de r\u00e9gression qui pr\u00e9voient la valeur d’une variable cible.…","og_url":"https:\/\/francestat.com\/index.php\/uniwin-gboost\/","og_site_name":"FRANCESTAT","article_modified_time":"2026-07-09T08:54:56+00:00","twitter_card":"summary_large_image","twitter_misc":{"Dur\u00e9e de lecture estim\u00e9e":"2 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/francestat.com\/index.php\/uniwin-gboost\/","url":"https:\/\/francestat.com\/index.php\/uniwin-gboost\/","name":"Uniwin - GBOOST - FRANCESTAT","isPartOf":{"@id":"https:\/\/francestat.com\/#website"},"datePublished":"2026-07-09T08:10:04+00:00","dateModified":"2026-07-09T08:54:56+00:00","breadcrumb":{"@id":"https:\/\/francestat.com\/index.php\/uniwin-gboost\/#breadcrumb"},"inLanguage":"fr-FR","potentialAction":[{"@type":"ReadAction","target":["https:\/\/francestat.com\/index.php\/uniwin-gboost\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/francestat.com\/index.php\/uniwin-gboost\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Accueil","item":"https:\/\/francestat.com\/"},{"@type":"ListItem","position":2,"name":"Uniwin – GBOOST"}]},{"@type":"WebSite","@id":"https:\/\/francestat.com\/#website","url":"https:\/\/francestat.com\/","name":"FRANCESTAT","description":"Logiciels, formations et services statistiques","publisher":{"@id":"https:\/\/francestat.com\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/francestat.com\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"fr-FR"},{"@type":"Organization","@id":"https:\/\/francestat.com\/#organization","name":"FRANCESTAT","url":"https:\/\/francestat.com\/","logo":{"@type":"ImageObject","inLanguage":"fr-FR","@id":"https:\/\/francestat.com\/#\/schema\/logo\/image\/","url":"https:\/\/francestat.com\/wp-content\/uploads\/2018\/05\/logo_horizontal_small_francestat.jpg","contentUrl":"https:\/\/francestat.com\/wp-content\/uploads\/2018\/05\/logo_horizontal_small_francestat.jpg","width":155,"height":51,"caption":"FRANCESTAT"},"image":{"@id":"https:\/\/francestat.com\/#\/schema\/logo\/image\/"}}]}},"_links":{"self":[{"href":"https:\/\/francestat.com\/index.php\/wp-json\/wp\/v2\/pages\/7352","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/francestat.com\/index.php\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/francestat.com\/index.php\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/francestat.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/francestat.com\/index.php\/wp-json\/wp\/v2\/comments?post=7352"}],"version-history":[{"count":14,"href":"https:\/\/francestat.com\/index.php\/wp-json\/wp\/v2\/pages\/7352\/revisions"}],"predecessor-version":[{"id":7372,"href":"https:\/\/francestat.com\/index.php\/wp-json\/wp\/v2\/pages\/7352\/revisions\/7372"}],"wp:attachment":[{"href":"https:\/\/francestat.com\/index.php\/wp-json\/wp\/v2\/media?parent=7352"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}