{"id":6799,"date":"2020-09-04T07:47:34","date_gmt":"2020-09-04T06:47:34","guid":{"rendered":"https:\/\/datascientest.com\/?p=6799"},"modified":"2023-11-09T14:12:32","modified_gmt":"2023-11-09T14:12:32","slug":"le-word-embedding","status":"publish","type":"post","link":"https:\/\/datascientest.com\/le-word-embedding","title":{"rendered":"Word Embedding : Un outil incontournable du NLP"},"content":{"rendered":"\t\t<div data-elementor-type=\"wp-post\" data-elementor-id=\"6799\" class=\"elementor elementor-6799\" 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-9842122 elementor-section-boxed elementor-section-height-default elementor-section-height-default\" data-id=\"9842122\" 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-cbcd81c\" data-id=\"cbcd81c\" 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-5719e5d elementor-widget elementor-widget-text-editor\" data-id=\"5719e5d\" 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<h2>Dans un <a href=\"https:\/\/datascientest.com\/introduction-au-nlp-natural-language-processing\" target=\"_blank\" rel=\"noopener\">article pr\u00e9cedent<\/a>, nous avons d\u00e9fini le NLP- Natural Language Processing. Dans cet article, nous allons nous int\u00e9resser \u00e0 l&rsquo;une de ses m\u00e9thodes principales, le word embedding. Le word embedding (plongement de mots) d\u00e9signe un ensemble de m\u00e9thode d\u2019apprentissage visant \u00e0 repr\u00e9senter les mots d\u2019un texte par des vecteurs de nombres r\u00e9els. Dans cet article, vous allez explorer trois grandes m\u00e9thodes pour repr\u00e9senter un mot sous forme d\u2019un vecteur :<\/h2><ul><li>M\u00e9thode 1 : One Hot Encoding<\/li><li>M\u00e9thode 2 : Image Embedding<\/li><li>M\u00e9thode 3 : Word Embedding<\/li><\/ul>\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-54ee5b8 elementor-section-boxed elementor-section-height-default elementor-section-height-default\" data-id=\"54ee5b8\" 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-baed581\" data-id=\"baed581\" data-element_type=\"column\">\n\t\t\t<div class=\"elementor-widget-wrap elementor-element-populated\">\n\t\t\t\t\t\t<div class=\"elementor-element elementor-element-6798d3b e-flex e-con-boxed e-con\" data-id=\"6798d3b\" 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-24e7bf1 elementor-widget__width-initial elementor-widget elementor-widget-image\" data-id=\"24e7bf1\" 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=\"800\" height=\"759\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/09\/Frise.png\" class=\"attachment-large size-large wp-image-7985\" alt=\"word embedding\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/09\/Frise.png 1778w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/09\/Frise-300x285.png 300w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/>\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\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-d465fb2 elementor-section-boxed elementor-section-height-default elementor-section-height-default\" data-id=\"d465fb2\" 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-655be8e\" data-id=\"655be8e\" 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-9a6c7f0 elementor-widget elementor-widget-heading\" data-id=\"9a6c7f0\" 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\">One hot encoding : Une m\u00e9thode qui transforme les lettres en valeurs num\u00e9riques<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-5bf581e elementor-widget elementor-widget-text-editor\" data-id=\"5bf581e\" 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\tLa repr\u00e9sentation vectorielle la plus classique des mots est le\u00a0one hot encoding. Une dimension est allou\u00e9e pour chaque mot du vocabulaire. Chaque mot du vocabulaire est repr\u00e9sent\u00e9 comme un vecteur binaire avec toutes ses valeurs nulles \u00e0 l&rsquo;exception de l&rsquo;index du mot.\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-5d700a3 elementor-widget elementor-widget-image\" data-id=\"5d700a3\" 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:\/\/medium.com\/@athif.shaffy\/one-hot-encoding-of-text-b69124bef0a7\">\n\t\t\t\t\t\t\t<img decoding=\"async\" width=\"800\" height=\"420\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/1-1024x537.png\" class=\"attachment-large size-large wp-image-6970\" alt=\"\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/1-1024x537.png 1024w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/1-300x157.png 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/1-768x403.png 768w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/1.png 1235w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/>\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\"><\/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-679dd0b elementor-widget elementor-widget-text-editor\" data-id=\"679dd0b\" 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>Pour illustrer l&rsquo;encodage, prenons la phrase suivante :<\/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-f34ae55 elementor-widget elementor-widget-text-editor\" data-id=\"f34ae55\" 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<pre><span style=\"color: #000000; font-family: 'Courier New'; font-size: 14px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: 400; white-space: pre-wrap;\">\"I think therefore I am\"<\/span><\/pre>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-fc070c2 elementor-widget elementor-widget-text-editor\" data-id=\"fc070c2\" 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>Le principe est d\u2019attribuer un indice \u00e0 chaque mot du vocabulaire. Le vocabulaire se pr\u00e9sente alors sous forme d\u2019un dictionnaire :<\/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-36d81e7 elementor-widget elementor-widget-text-editor\" data-id=\"36d81e7\" 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<pre style=\"padding-left: 280px;\"><span id=\"docs-internal-guid-70e932f0-7fff-b019-7e84-99f159a91fe9\" style=\"font-weight: normal;\"><span style=\"font-size: 10.5pt; font-family: 'Courier New'; color: #000000; font-weight: 400; font-style: normal; font-variant: normal; white-space: pre-wrap;\">{'am': 0, 'i': 1, 'therefore': 2, 'think': 3}<\/span><\/span><\/pre>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-cf6cd00 elementor-widget elementor-widget-text-editor\" data-id=\"cf6cd00\" 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;\">Ici, <\/span><b>le vocabulaire est compos\u00e9 les 4 mots uniques de notre corpus<\/b><span style=\"font-weight: 400;\">. La m\u00e9thode consiste alors \u00e0 repr\u00e9senter le mot du vocabulaire sous forme de vecteur de dimension 4 (taille du vocabulaire) qui a toutes ses valeurs nulles \u00e0 l&rsquo;exception de l&rsquo;index du mot.<\/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-5a0e5b4 elementor-widget elementor-widget-text-editor\" data-id=\"5a0e5b4\" 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<pre><span id=\"docs-internal-guid-2b70bf31-7fff-f276-db3b-9d9b9b3829b7\" style=\"font-weight: normal;\"><span style=\"font-size: 11pt; font-family: 'Cambria Math', serif; color: #000000; background-color: transparent; font-weight: 400; font-variant: normal; white-space: pre-wrap;\">Onehot(<\/span><span style=\"font-size: 11pt; font-family: 'Cambria Math', serif; color: #000000; background-color: transparent; font-weight: 400; font-variant: normal; white-space: pre-wrap;\">'i')<\/span><span style=\"font-size: 11pt; font-family: 'Cambria Math', serif; color: #000000; background-color: transparent; font-weight: 400; font-variant: normal; white-space: pre-wrap;\">=<\/span><span style=\"font-size: 11pt; font-family: 'Cambria Math', serif; color: #000000; background-color: transparent; font-weight: 400; font-variant: normal; white-space: pre-wrap;\">0<\/span> <span style=\"font-size: 11pt; font-family: 'Cambria Math', serif; color: #000000; background-color: transparent; font-weight: 400; font-variant: normal; white-space: pre-wrap;\">1<\/span> <span style=\"font-size: 11pt; font-family: 'Cambria Math', serif; color: #000000; background-color: transparent; font-weight: 400; font-variant: normal; white-space: pre-wrap;\">0<\/span> <span style=\"font-size: 11pt; font-family: 'Cambria Math', serif; color: #000000; background-color: transparent; font-weight: 400; font-variant: normal; white-space: pre-wrap;\">0<\/span> <\/span><\/pre>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-da83fe9 elementor-widget elementor-widget-text-editor\" data-id=\"da83fe9\" 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;\">Avec cette repr\u00e9sentation, <\/span><b>tous les mots ont la m\u00eame distance et la m\u00eame similitude<\/b><span style=\"font-weight: 400;\">. L&rsquo;encodage\u00a0<\/span><b>one hot<\/b><span style=\"font-weight: 400;\">\u00a0n&rsquo;apporte donc qu&rsquo;une information selon laquelle un <\/span><b>mot est diff\u00e9rent d&rsquo;un autre<\/b><span style=\"font-weight: 400;\">.<\/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-c3e4687 elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"c3e4687\" 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\">Apprendre \u00e0 ma\u00eetriser le one hot encoding<\/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<section class=\"elementor-section elementor-top-section elementor-element elementor-element-a57daa8 elementor-section-boxed elementor-section-height-default elementor-section-height-default\" data-id=\"a57daa8\" 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-ab95c8f\" data-id=\"ab95c8f\" 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-f0f7a6e elementor-widget elementor-widget-heading\" data-id=\"f0f7a6e\" 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 transformer des images en valeurs num\u00e9riques \u00e0 l'aide de l'image Embedding ?<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-c3491bd elementor-widget elementor-widget-text-editor\" data-id=\"c3491bd\" 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 dimension d&rsquo;une image est d\u00e9finie par son nombre de pixel. Dans un article pr\u00e9c\u00e9dent sur les\u00a0<a href=\"https:\/\/datascientest.com\/convolutional-neural-network\">Convolutionnal neural network<\/a>, nous utilisons des <\/span><b>convolutions pour l&rsquo;extraction de caract\u00e9ristiques plus pertinentes<\/b> <b>que la valeur des pixels<\/b><span style=\"font-weight: 400;\">. Par exemple, les noyaux servant \u00e0 d\u00e9tecter les bords pourraient \u00eatre utiles pour classifier certaines formes g\u00e9om\u00e9triques.<\/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-0d9bcd5 elementor-widget elementor-widget-image\" data-id=\"0d9bcd5\" 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:\/\/towardsdatascience.com\/a-comprehensive-guide-to-convolutional-neural-networks-the-eli5-way-3bd2b1164a53\">\n\t\t\t\t\t\t\t<img decoding=\"async\" width=\"800\" height=\"275\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/2-1-1024x352.png\" class=\"attachment-large size-large wp-image-6971\" alt=\"\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/2-1-1024x352.png 1024w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/2-1-300x103.png 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/2-1-768x264.png 768w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/2-1.png 1337w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/>\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\">Convolution Neuronal 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-3bc0bbb elementor-widget elementor-widget-text-editor\" data-id=\"3bc0bbb\" 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;\">L&rsquo;image embedding\u00a0est capable en utilisant les convolutions\/pooling d&rsquo;encoder l&rsquo;image et\u00a0d&rsquo;extraire des <\/span><b>relations de lin\u00e9arit\u00e9 entre des propri\u00e9t\u00e9s (style artistique, forme sp\u00e9cifique&#8230;) et les nouvelles caract\u00e9ristiques de notre image.<\/b><\/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-2f351f0 elementor-widget elementor-widget-image\" data-id=\"2f351f0\" 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.oreilly.com\/library\/view\/practical-deep-learning\/9781492034858\/ch04.html\">\n\t\t\t\t\t\t\t<img decoding=\"async\" width=\"300\" height=\"296\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/3-1-300x296.png\" class=\"attachment-medium size-medium wp-image-6972\" alt=\"\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/3-1-300x296.png 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/3-1-1024x1009.png 1024w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/3-1-768x757.png 768w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/3-1.png 1440w\" sizes=\"(max-width: 300px) 100vw, 300px\" \/>\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\">ACP sur l'embedding des images<\/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\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-76a7d3d elementor-section-boxed elementor-section-height-default elementor-section-height-default\" data-id=\"76a7d3d\" 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-54f9b2a\" data-id=\"54f9b2a\" 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-3a83ce6 elementor-widget elementor-widget-heading\" data-id=\"3a83ce6\" 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\">D\u00e9finition du Word Embedding <\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-7b4d7c8 elementor-widget elementor-widget-text-editor\" data-id=\"7b4d7c8\" 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 analogie<\/span><b>, le word embedding est capable en r\u00e9duisant la dimension de capturer le contexte, la similarit\u00e9 s\u00e9mantique et syntaxique (genre, synonymes, &#8230;) d&rsquo;un mot<\/b><span style=\"font-weight: 400;\">. Par exemple, on pourrait s&rsquo;attendre \u00e0 ce que les mots \u00ab chien \u00bb et \u00ab chat \u00bb soient repr\u00e9sent\u00e9s par des vecteurs relativement peu distants dans l&rsquo;espace vectoriel o\u00f9 sont d\u00e9finis ces vecteurs.<\/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-ebd25e4 elementor-widget elementor-widget-text-editor\" data-id=\"ebd25e4\" 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 pour les images, nous souhaitons que \u00e7a soit <\/span><b>le mod\u00e8le qui choisisse les caract\u00e9ristiques les plus pertinentes repr\u00e9sentant le mot<\/b><span style=\"font-weight: 400;\">. Par exemple, la caract\u00e9ristique \u00ab\u00a0\u00eatre vivant\u00a0\u00bb pourrait \u00eatre int\u00e9ressante pour diff\u00e9rencier \u00ab\u00a0chien\u00a0\u00bb et \u00ab\u00a0ordinateur\u00a0\u00bb, et rapprocher \u00ab\u00a0chien\u00a0\u00bb et \u00ab\u00a0chat\u00a0\u00bb.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-4e897e5 e-flex e-con-boxed e-con\" data-id=\"4e897e5\" 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-dd41eea elementor-widget__width-initial elementor-widget elementor-widget-video\" data-id=\"dd41eea\" data-element_type=\"widget\" data-settings=\"{&quot;youtube_url&quot;:&quot;https:\\\/\\\/youtu.be\\\/jC4UnPf7oZo&quot;,&quot;video_type&quot;:&quot;youtube&quot;,&quot;controls&quot;:&quot;yes&quot;}\" data-widget_type=\"video.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-video .elementor-widget-container{overflow:hidden;transform:translateZ(0)}.elementor-widget-video .elementor-wrapper{aspect-ratio:var(--video-aspect-ratio)}.elementor-widget-video .elementor-wrapper iframe,.elementor-widget-video .elementor-wrapper video{height:100%;width:100%;display:flex;border:none;background-color:#000}@supports not (aspect-ratio:1\/1){.elementor-widget-video .elementor-wrapper{position:relative;overflow:hidden;height:0;padding-bottom:calc(100% \/ var(--video-aspect-ratio))}.elementor-widget-video .elementor-wrapper iframe,.elementor-widget-video .elementor-wrapper video{position:absolute;top:0;right:0;bottom:0;left:0}}.elementor-widget-video .elementor-open-inline .elementor-custom-embed-image-overlay{position:absolute;top:0;right:0;bottom:0;left:0;background-size:cover;background-position:50%}.elementor-widget-video .elementor-custom-embed-image-overlay{cursor:pointer;text-align:center}.elementor-widget-video .elementor-custom-embed-image-overlay:hover .elementor-custom-embed-play i{opacity:1}.elementor-widget-video .elementor-custom-embed-image-overlay img{display:block;width:100%;aspect-ratio:var(--video-aspect-ratio);-o-object-fit:cover;object-fit:cover;-o-object-position:center center;object-position:center center}@supports not (aspect-ratio:1\/1){.elementor-widget-video .elementor-custom-embed-image-overlay{position:relative;overflow:hidden;height:0;padding-bottom:calc(100% \/ var(--video-aspect-ratio))}.elementor-widget-video .elementor-custom-embed-image-overlay img{position:absolute;top:0;right:0;bottom:0;left:0}}.elementor-widget-video .e-hosted-video .elementor-video{-o-object-fit:cover;object-fit:cover}.e-con-inner>.elementor-widget-video,.e-con>.elementor-widget-video{width:var(--container-widget-width);--flex-grow:var(--container-widget-flex-grow)}<\/style>\t\t<div class=\"elementor-wrapper elementor-open-inline\">\n\t\t\t<div class=\"elementor-video\"><\/div>\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<\/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-f0d238c elementor-section-boxed elementor-section-height-default elementor-section-height-default\" data-id=\"f0d238c\" 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-08fc990\" data-id=\"08fc990\" 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-c5e091f elementor-widget elementor-widget-heading\" data-id=\"c5e091f\" 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\">Matrice d\u2019embedding<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-c9bc700 elementor-widget elementor-widget-text-editor\" data-id=\"c9bc700\" 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 d\u2019embedding qui g\u00e9n\u00e9ralement utilis\u00e9 pour r\u00e9duire la dimension d\u2019un vecteur est d\u2019utiliser le r\u00e9sultat que retourne une dense layer comme embedding, c&rsquo;est \u00e0 dire de multiplier une matrice d\u2019embedding\u00a0W\u00a0par la repr\u00e9sentation \u00ab\u00a0one hot\u00a0\u00bb du mot :<\/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-1361ac3 elementor-widget elementor-widget-image\" data-id=\"1361ac3\" 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=\"332\" height=\"162\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/2-3.png\" class=\"attachment-large size-large wp-image-6976\" alt=\"\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/2-3.png 332w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/2-3-300x146.png 300w\" sizes=\"(max-width: 332px) 100vw, 332px\" \/>\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-87fb68b elementor-widget elementor-widget-text-editor\" data-id=\"87fb68b\" 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;\">Sous forme vectorielle :<\/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-6323cbe elementor-widget elementor-widget-image\" data-id=\"6323cbe\" 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=\"313\" height=\"53\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/5.png\" class=\"attachment-large size-large wp-image-6977\" alt=\"\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/5.png 313w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/5-300x51.png 300w\" sizes=\"(max-width: 313px) 100vw, 313px\" \/>\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-89267cd elementor-widget elementor-widget-image\" data-id=\"89267cd\" 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\t<a href=\"https:\/\/medium.com\/datadriveninvestor\/deep-learning-autoencoders-db265359943e\">\n\t\t\t\t\t\t\t<img decoding=\"async\" width=\"254\" height=\"300\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/4-254x300.png\" class=\"attachment-medium size-medium wp-image-6978\" alt=\"\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/4-254x300.png 254w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/4.png 454w\" sizes=\"(max-width: 254px) 100vw, 254px\" \/>\t\t\t\t\t\t\t\t<\/a>\n\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-d78e207 elementor-widget elementor-widget-text-editor\" data-id=\"d78e207\" 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;\">Maintenant que l\u2019on a d\u00e9fini la m\u00e9thode de r\u00e9duction de dimension (de compression de l\u2019information), comment est-ce qu\u2019on entra\u00eene la matrice d\u2019embedding <\/span><b>W<\/b><span style=\"font-weight: 400;\">\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-8e52cdb elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"8e52cdb\" 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\">Suivre une formation en Data Science<\/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<section class=\"elementor-section elementor-top-section elementor-element elementor-element-5536c81 elementor-section-boxed elementor-section-height-default elementor-section-height-default\" data-id=\"5536c81\" 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-0c1809d\" data-id=\"0c1809d\" 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-32d73e1 elementor-widget elementor-widget-heading\" data-id=\"32d73e1\" 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\">Entra\u00eenement de la matrice d\u2019embedding \u2013 Probl\u00e8me de classification<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-c3177a8 elementor-widget elementor-widget-text-editor\" data-id=\"c3177a8\" 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;\">Une premi\u00e8re fa\u00e7on de trouver la<\/span><b> matrice d&#8217;embedding est de l\u2019entra\u00eener sur un probl\u00e8me supervis\u00e9<\/b><span style=\"font-weight: 400;\">. De ce fait, nous allons utiliser <\/span><b>cette repr\u00e9sentation des mots pour r\u00e9soudre un probl\u00e8me de sentiment analysis<\/b><span style=\"font-weight: 400;\"> sur les reviews IMBD. Le jeu de donn\u00e9es comporte 25000 critiques de films (plus d&rsquo;informations sur <a href=\"https:\/\/ai.stanford.edu\/~amaas\/data\/sentiment\/\">le jeu de donn\u00e9es choisi<\/a>).<\/span><\/p><p><span style=\"font-weight: 400;\">Il est alors possible d&rsquo;<\/span><b>entra\u00eener la matrice\u00a0W\u00a0du word embedding en m\u00eame temps que l&rsquo;entra\u00eenement du probl\u00e8me de classification.<\/b><span style=\"font-weight: 400;\"><br \/><\/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-4a7203c elementor-widget elementor-widget-text-editor\" data-id=\"4a7203c\" 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; font-family: Calibri, sans-serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: bold; text-align: justify; white-space: pre-wrap;\">Mise en forme des donn\u00e9es\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-13cd460 elementor-widget elementor-widget-text-editor\" data-id=\"13cd460\" 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>Pour mettre en forme les donn\u00e9es, il est n\u00e9cessaire de vectoriser le corpus de texte, en transformant chaque review en une s\u00e9quence d&rsquo;entiers (chaque entier \u00e9tant l&rsquo;index d&rsquo;un mot)\u00a0:<\/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-5ec18b6 elementor-widget elementor-widget-text-editor\" data-id=\"5ec18b6\" 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 dir=\"ltr\" style=\"line-height: 1.2; text-align: justify; background-color: #ffffff; padding: 0pt 0pt 12pt 0pt; margin: 12pt 24pt 0pt 24pt;\"><span style=\"font-size: 10.5pt; font-family: 'Courier New'; color: #000000; background-color: #ffffff; font-weight: 400; font-style: normal; font-variant: normal; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;\">tokenizer.texts_to_sequences([&lsquo;hello my dear readers&rsquo;]) <\/span><span style=\"color: rgb(0, 0, 0); font-family: &quot;Courier New&quot;; font-size: 10.5pt; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: 400; white-space: pre-wrap;\">-&gt; [[4422, 10, 2974, 6117]]<\/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-dc7e0ac elementor-widget elementor-widget-image\" data-id=\"dc7e0ac\" 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=\"800\" height=\"610\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/2-2.png\" class=\"attachment-large size-large wp-image-6973\" alt=\"\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/2-2.png 940w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/2-2-300x229.png 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2020\/08\/2-2-768x586.png 768w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/>\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-8cc2d86 elementor-widget elementor-widget-text-editor\" data-id=\"8cc2d86\" 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;\">Notre mod\u00e8le sera un simple\u00a0bag of word avec une m\u00e9thode d\u2019embedding\u00a0:<\/span><\/p><ol><li><span style=\"font-weight: 400;\">La couche\u00a0Embedding\u00a0va transformer chaque index de mots en <\/span><b>vecteur d&#8217;embedding<\/b><span style=\"font-weight: 400;\">. La matrice\u00a0W\u00a0de l&#8217;embedding sera apprise au fur et \u00e0 mesure que le mod\u00e8le s&rsquo;entra\u00eene. Les dimensions r\u00e9sultantes sont : (lot, s\u00e9quence, embedding).<\/span><\/li><li>\u00a0<span style=\"font-weight: 400;\">Ensuite, la couche GlobalAveragePooling1D renvoie un vecteur de sortie de longueur fixe pour chaque exemple en faisant la moyenne sur la dimension de s\u00e9quence. Cette transformation consiste \u00e0 faire un <\/span><b>bag of word<\/b><span style=\"font-weight: 400;\"> et permet au mod\u00e8le de g\u00e9rer des entr\u00e9es de longueur variable.<\/span><\/li><li>\u00a0<span style=\"font-weight: 400;\">Enfin, comme nous sommes face \u00e0 un probl\u00e8me de classification (review positive ou n\u00e9gative), il est n\u00e9cessaire d\u2019<\/span><b>ajouter des couches\u00a0Dense<\/b><span style=\"font-weight: 400;\">\u00a0pour classifier le sentiment de la review.<\/span><\/li><\/ol>\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-f3b5cb1 elementor-section-boxed elementor-section-height-default elementor-section-height-default\" data-id=\"f3b5cb1\" 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-b2a4f49\" data-id=\"b2a4f49\" 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-b2418b0 elementor-widget elementor-widget-heading\" data-id=\"b2418b0\" 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\">Mise en pratique<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-1d9a9ec elementor-widget elementor-widget-shortcode\" data-id=\"1d9a9ec\" data-element_type=\"widget\" data-widget_type=\"shortcode.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<div class=\"elementor-shortcode\"><script src=\"https:\/\/gist.github.com\/15boehler\/4437a45f7c2679b319dca43be6dcee1b.js\"><\/script><\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-063fc81 elementor-widget elementor-widget-text-editor\" data-id=\"063fc81\" 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;\">De cette fa\u00e7on, nous avons atteint sur le jeu de donn\u00e9es de test une pr\u00e9cision de 0.88. Ce score est difficilement atteignable \u00e0 l\u2019aide de m\u00e9thodes classiques. Par ailleurs, nous aurions aussi pu utiliser une <\/span><b>m\u00e9thode tf-idf<\/b><span style=\"font-weight: 400;\"> (\u00e0 la place du bag of word) pour transformer notre s\u00e9quence en vecteur.<\/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-8648882 elementor-widget elementor-widget-heading\" data-id=\"8648882\" 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\">Pour r\u00e9sumer :<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-7dc7a1f elementor-widget elementor-widget-text-editor\" data-id=\"7dc7a1f\" 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;\">Un mod\u00e8le de ce type donne en g\u00e9n\u00e9ral de meilleur r\u00e9sultat que les approches traditionnelles. Il permet \u00e9galement de r\u00e9duire la dimension du probl\u00e8me et par cons\u00e9quent la t\u00e2che d&rsquo;apprentissage. N\u00e9anmoins, comme nous ne donnons qu\u2019une information sur le sentiment, le word embedding ne capture que difficilement d\u2019autre relation que \u00e7a.<\/span><\/p><p><span style=\"font-weight: 400;\">Dans un prochain article, nous allons entra\u00eener la matrice d\u2019embedding de mani\u00e8re non-supervis\u00e9 \u00e0 l&rsquo;aide du c\u00e9l\u00e8bre algorithme word2vec. Comme l\u2019entra\u00eenement se fait de mani\u00e8re non supervis\u00e9e, nous allons \u00e9viter le biais li\u00e9 \u00e0 la r\u00e9solution d\u2019un probl\u00e8me supervis\u00e9e.<\/span><span style=\"font-weight: 400;\"><br \/><\/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-3965d73 elementor-section-boxed elementor-section-height-default elementor-section-height-default\" data-id=\"3965d73\" 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-9f1c638\" data-id=\"9f1c638\" 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-b5612f4 elementor-widget elementor-widget-text-editor\" data-id=\"b5612f4\" 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>Cet article vous a plu ? Vous souhaitez aller plus loin dans votre apprentissage de la data science ?\u00a0<\/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-fb42213 elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"fb42213\" 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\u00e9marrer une formation en Data Science<\/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<div class=\"elementor-element elementor-element-3b50e6e e-flex e-con-boxed e-con\" data-id=\"3b50e6e\" 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-2d89acc elementor-widget elementor-widget-html\" data-id=\"2d89acc\" 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    \"@type\": \"Question\",\r\n    \"name\": \"One hot encoding : Une m\u00e9thode qui transforme les lettres en valeurs num\u00e9riques\",\r\n    \"acceptedAnswer\": {\r\n      \"@type\": \"Answer\",\r\n      \"text\": \"La repr\u00e9sentation vectorielle la plus classique des mots est le one hot encoding. Une dimension est allou\u00e9e pour chaque mot du vocabulaire. Chaque mot du vocabulaire est repr\u00e9sent\u00e9 comme un vecteur binaire avec toutes ses valeurs nulles \u00e0 l'exception de l'index du mot.\"\r\n    }\r\n  },{\r\n    \"@type\": \"Question\",\r\n    \"name\": \"Comment transformer des images en valeurs num\u00e9riques \u00e0 l'aide de l'image Embedding ?\",\r\n    \"acceptedAnswer\": {\r\n      \"@type\": \"Answer\",\r\n      \"text\": \"La dimension d'une image est d\u00e9finie par son nombre de pixel. Dans un article pr\u00e9c\u00e9dent sur les Convolutionnal neural network, nous utilisons des convolutions pour l'extraction de caract\u00e9ristiques plus pertinentes que la valeur des pixels. Par exemple, les noyaux servant \u00e0 d\u00e9tecter les bords pourraient \u00eatre utiles pour classifier certaines formes g\u00e9om\u00e9triques.\"\r\n    }\r\n  },{\r\n    \"@type\": \"Question\",\r\n    \"name\": \"Entra\u00eenement de la matrice d\u2019embedding \u2013 Probl\u00e8me de classification\",\r\n    \"acceptedAnswer\": {\r\n      \"@type\": \"Answer\",\r\n      \"text\": \"Une premi\u00e8re fa\u00e7on de trouver la matrice d'embedding est de l\u2019entra\u00eener sur un probl\u00e8me supervis\u00e9. De ce fait, nous allons utiliser cette repr\u00e9sentation des mots pour r\u00e9soudre un probl\u00e8me de sentiment analysis sur les reviews IMBD. Le jeu de donn\u00e9es comporte 25000 critiques de films (plus d'informations sur le jeu de donn\u00e9es choisi). Il est alors possible d'entra\u00eener la matrice W du word embedding en m\u00eame temps que l'entra\u00eenement du probl\u00e8me de classification.\"\r\n    }\r\n  }]\r\n}\r\n<\/script>\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>Dans un article pr\u00e9cedent, nous avons d\u00e9fini le NLP- Natural Language Processing. Dans cet article, nous allons nous int\u00e9resser \u00e0 l&rsquo;une de ses m\u00e9thodes principales, le word embedding. Le word embedding (plongement de mots) d\u00e9signe un ensemble de m\u00e9thode d\u2019apprentissage visant \u00e0 repr\u00e9senter les mots d\u2019un texte par des vecteurs de nombres r\u00e9els. Dans cet [&hellip;]<\/p>\n","protected":false},"author":78,"featured_media":199148,"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>Word Embedding : Un outil incontournable du NLP<\/title>\n<meta name=\"description\" content=\"D\u00e9couvrez le word embedding, une des m\u00e9thodes principales utilis\u00e9es pour le NLP- Natural Language Processing. On vous explique !\" \/>\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\/le-word-embedding\" \/>\n<meta property=\"og:locale\" content=\"fr_FR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Word Embedding : Un outil incontournable du NLP\" \/>\n<meta property=\"og:url\" content=\"https:\/\/datascientest.com\/le-word-embedding\" \/>\n<meta property=\"og:site_name\" content=\"Formation Data Science | DataScientest.com\" \/>\n<meta property=\"article:published_time\" content=\"2020-09-04T06:47:34+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2023-11-09T14:12:32+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/09\/word-embedding-768x384-1.png\" \/>\n\t<meta property=\"og:image:width\" content=\"768\" \/>\n\t<meta property=\"og:image:height\" content=\"384\" \/>\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=\"7 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/datascientest.com\/le-word-embedding\",\"url\":\"https:\/\/datascientest.com\/le-word-embedding\",\"name\":\"Word Embedding : Un outil incontournable du NLP\",\"isPartOf\":{\"@id\":\"https:\/\/datascientest.com\/#website\"},\"datePublished\":\"2020-09-04T06:47:34+00:00\",\"dateModified\":\"2023-11-09T14:12:32+00:00\",\"author\":{\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/7af2361c48e24e79a8aa2360e6e74925\"},\"description\":\"D\u00e9couvrez le word embedding, une des m\u00e9thodes principales utilis\u00e9es pour le NLP- Natural Language Processing. On vous explique !\",\"inLanguage\":\"fr-FR\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/datascientest.com\/le-word-embedding\"]}]},{\"@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":"Word Embedding : Un outil incontournable du NLP","description":"D\u00e9couvrez le word embedding, une des m\u00e9thodes principales utilis\u00e9es pour le NLP- Natural Language Processing. On vous explique !","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\/le-word-embedding","og_locale":"fr_FR","og_type":"article","og_title":"Word Embedding : Un outil incontournable du NLP","og_url":"https:\/\/datascientest.com\/le-word-embedding","og_site_name":"Formation Data Science | DataScientest.com","article_published_time":"2020-09-04T06:47:34+00:00","article_modified_time":"2023-11-09T14:12:32+00:00","og_image":[{"width":768,"height":384,"url":"https:\/\/datascientest.com\/wp-content\/uploads\/2020\/09\/word-embedding-768x384-1.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":"7 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/datascientest.com\/le-word-embedding","url":"https:\/\/datascientest.com\/le-word-embedding","name":"Word Embedding : Un outil incontournable du NLP","isPartOf":{"@id":"https:\/\/datascientest.com\/#website"},"datePublished":"2020-09-04T06:47:34+00:00","dateModified":"2023-11-09T14:12:32+00:00","author":{"@id":"https:\/\/datascientest.com\/#\/schema\/person\/7af2361c48e24e79a8aa2360e6e74925"},"description":"D\u00e9couvrez le word embedding, une des m\u00e9thodes principales utilis\u00e9es pour le NLP- Natural Language Processing. On vous explique !","inLanguage":"fr-FR","potentialAction":[{"@type":"ReadAction","target":["https:\/\/datascientest.com\/le-word-embedding"]}]},{"@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\/6799"}],"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=6799"}],"version-history":[{"count":0,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts\/6799\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/media\/199148"}],"wp:attachment":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/media?parent=6799"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/categories?post=6799"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/tags?post=6799"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}