{"id":1769,"date":"2022-11-23T10:17:55","date_gmt":"2022-11-23T09:17:55","guid":{"rendered":"https:\/\/enpc.ergeais.com\/?p=1769"},"modified":"2025-07-29T14:29:45","modified_gmt":"2025-07-29T12:29:45","slug":"artificial-intelligence-a-scientific-and-technological-revolution","status":"publish","type":"post","link":"https:\/\/ingenius.ecoledesponts.fr\/en\/articles\/artificial-intelligence-a-scientific-and-technological-revolution\/","title":{"rendered":"Artificial Intelligence: a scientific and technological revolution"},"content":{"rendered":"\n\n\n<figure class=\"wp-block-image alignwide size-large is-style-default\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"589\" src=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_01_Grande_illu-1024x589.jpg\" alt=\"\" class=\"wp-image-863\" srcset=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_01_Grande_illu-1024x589.jpg 1024w, https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_01_Grande_illu-300x173.jpg 300w, https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_01_Grande_illu-768x442.jpg 768w, https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_01_Grande_illu-1920x1104.jpg 1920w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><figcaption class=\"wp-element-caption\">&#8220;Artificial Intelligence, Thinking Machines and the Future of Humanity&#8221; \u00a9 Gerd Leonhard, CC-BY-SA (source : flickr)<\/figcaption><\/figure>\n\n\n\n<p>Before we can speak of cybernetics, Artificial Intelligence (AI) requires human intelligence, and that is what drives the research teams at \u00c9cole nationale des ponts et chauss\u00e9es. Like his peers who are working to push the limits of AI in France and around the world, Axel Parmentier, researcher at <a href=\"https:\/\/cermics-lab.enpc.fr\/\" target=\"_blank\" rel=\"noreferrer noopener\">CERMICS <\/a>(Center of Teaching and Research in Mathematics and Scientific Calculation), is one of the players in this scientific and technological revolution that is shaking up countless activities: from the digital economy to the autonomous vehicle, including the optimization of industrial processes. \u201cWe have inherited more than a century of tradition and very strong skills in applied mathematics and at the same time, we are the actors of a scientific and technological discipline that is buzzing,&#8221; explains the young researcher.<\/p>\n\n\n\n<p>Artificial intelligence today covers three essential types of applications:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>&nbsp;The first is data-mining, describing gigantic sets of data in order to extract a relevant summary. This applies to the bulk of applications that are now mature in the traditional economy,\u201d says Axel Parmentier ;<\/li>\n<\/ul>\n\n\n\n<p><\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>The second is the use of data to make predictions through supervised learning. It is notably a subject of expertise of the IMAGINE team (at the Gaspard-Monge computer laboratory), particularly in the field of computer vision ;<\/li>\n<\/ul>\n\n\n\n<p><\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>The third is decision-making support, which will make it possible to recommend the best possible decision on complex problems using data-driven optimization models, and on which the CERMICS optimization team is working in particular.<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>Operational research, a formidable decision support tool<\/strong><\/h2>\n\n\n\n<p>In this large AI family, Axel Parmentier, is an expert in operational research. This discipline combines applied mathematics and computer science to help industrialists allocate their resources more effectively. For example, it provides tools for planning a supply chain or managing energy networks (smart grids).<\/p>\n\n\n\n<p>It is implemented within the school through a chair with Air France inaugurated at the end of 2016, entitled \u201c<a href=\"https:\/\/ecoledesponts.fr\/intelligence-artificielle-pour-laerien\" data-type=\"URL\" data-id=\"https:\/\/ecoledesponts.fr\/intelligence-artificielle-pour-laerien\" target=\"_blank\" rel=\"noreferrer noopener\">Artificial intelligence for the Air Transport industry<\/a>\u201d, the first in France to focus on the interface between these disciplines. One of the creations by the research unit, along with the airline&#8217;s R&amp;D teams, is an algorithm that builds flight sequences for aircraft and crews and enables these flights to be operated at lower cost. This is a tricky issue for commercial software.On this type of problem, it is not uncommon for an algorithm to be able to solve problems for 500 flights in a few seconds, problems for 600 flights in a few hours, but be totally unable to solve problems for 700 flights<em>. <\/em>\u201cThis barrier is explained not by reasons of computing power, but by a wall of theoretical complexity\u201d explains Axel Parmentier. Therefore, mathematical models and algorithms have to be defined to find solutions, which makes great thesis topics for our laboratory!\u201d.&nbsp;<\/p>\n\n\n\n<h3 class=\"wp-block-heading has-medium-grey-color has-text-color\"><strong>Frank-Wolfe Algorithm<\/strong><\/h3>\n\n\n\n<figure class=\"wp-block-image aligncenter size-large is-resized\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"904\" src=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_02_Franck-Wolfe-1024x904.jpg\" alt=\"\" class=\"wp-image-737\" style=\"width:414px;height:365px\" srcset=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_02_Franck-Wolfe-1024x904.jpg 1024w, https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_02_Franck-Wolfe-300x265.jpg 300w, https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_02_Franck-Wolfe-768x678.jpg 768w, https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_02_Franck-Wolfe.jpg 1413w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><figcaption class=\"wp-element-caption\"><strong>In operational research, the aim is to find the best decision; in machine learning, it is a matter of finding the model that makes the best predictions. Mathematically, this amounts to finding the minimum of a function, here the point at the bottom of the blue sphere. Many efficient algorithms, such as the Franck-Wolfe algorithm illustrated here, are based on geometry. &#8220;Frank-Wolfe Algorithm&#8221; \u00a9 Stephanie Stutz, CC-BY-SA (source: Wikimedia Commons).<\/strong><\/figcaption><\/figure>\n\n\n\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>Significant progress thanks to increasingly efficient algorithms<\/strong><\/h2>\n\n\n\n<p>AI has been around for decades, but it has recently seen significant advances in a growing number of sectors. Why? \u201cThese advances are made possible by the increase in computer power or the availability of big data, but the key lies in the development of increasingly high-performance algorithms, i.e. in the work of researchers such as those in our laboratory\u201d says Axel Parmentier.<\/p>\n\n\n\n<p>The algorithm is, in his words, \u201cthe computer equivalent of a pastry chef&#8217;s recipe\u201d which, through a succession of elementary mathematical operations, allows the computer to perform tasks such as making decisions or predictions. A line of code in software is a basic operation. This is what researchers do on a daily basis! Take the example of the AlphaGo Zero algorithm by the company DeepMind which beat the human brain in 2015 in the game of Go, which is no more than a sequence of predictions and decisions.<\/p>\n\n\n\n<p>\u201cAn enormous amount of progress has been made in algorithms in recent decades. Take the example of deep learning: 20 years ago, the models didn\u2019t work very well, partly because there was a lack of data to exploit, but mainly because proper neural network architectures were not yet available. For example, running the 1990 integer linear programming algorithms (a flagship tool in operations research) with today&#8217;s computers can increase speed by hundreds of thousands of times over the computers of the day. With the latest generation of algorithms, the progress is by thousands of billions of times! Advances in learning and computer vision algorithms have led to advances in automatic recognition of objects in video sequences, a technology used in the autonomous car.<\/p>\n\n\n\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>A multitude of unsuspected applications<\/strong><\/h2>\n\n\n\n<p>\u00c9cole nationale des ponts et chauss\u00e9es contributes to the progress of research by developing new models and improving the understanding of those produced by the scientific community. Its industrial partners, such as Air France, Renault and Total, are looking for algorithms that can solve their problems, but above all the scientific guarantee of the result. \u201cBefore validating algorithms that enable the running of a power plant or the movement of a car, we must be certain of the quality of the results they produce, which requires an understanding of their mathematical properties,\u201d explains the researcher. \u201cThis is the beauty of mathematics: you model a concrete problem in the form of an abstract mathematical problem, you build a nice theory to solve this problem and you obtain efficient algorithms that find a multitude of unsuspected applications\u201d. And to quote linear integer programming, which makes it possible to model a wide variety of problems. Today, it has applications ranging from the placement of advertising spots on a television channel to airline company schedules, including the optimal management of hydroelectric dam valleys, or the organization of production for an industrial company.<\/p>\n\n\n\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>Cross-research in the heart of the laboratories<\/strong><strong><\/strong><\/h2>\n\n\n\n<p>The needs of the economy and scientific progress even lead to the mixing of disciplines within its laboratories. This is the case, for example, between CERMICS and IMAGINE, which traditionally operate on distinct fields: probability, numerical analysis and optimization for the former, statistics, prediction, artificial vision and machine learning for the latter. The evolution of technology, as in the reinforcement learning popularized by AlphaGo, blurs the boundaries between optimization and machine learning. \u201cThis encourages crossed research between our teams. This work is applied by our industrial partner Air France to predictive aircraft maintenance, i.e. the ability to use the data fed back by the aircraft to anticipate accurately when a particular part of an aircraft is likely to malfunction and thus avoid breakdowns\u201d, notes Axel Parmentier.By combining the fields of prediction and decision making, their common goal is to advance practical and theoretical knowledge.<\/p>\n\n\n\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>Towards a quest for excellence<\/strong><\/h2>\n\n\n\n<p>\u201cIn the age of artificial intelligence, the potential applications of applied mathematics are almost infinite, but our top priority remains academic excellence,\u201d says Axel Parmentier. As in other fields of research, this quest for excellence is also measured by regular publications in top-ranking scientific journals, frequent participation in conferences that set the pace in this field, cooperation with European or American universities or close collaboration with institutes such as <a href=\"https:\/\/www.inria.fr\/fr\" target=\"_blank\" rel=\"noreferrer noopener\">Inria<\/a> (National Institute for Research in Digital Science and Technology). One of the CERMICS teams has thus developed first-rate expertise with this institute in the field of molecular simulation, which should become a major area of progress in the coming years for the chemical and pharmaceutical industries.<\/p>\n\n\n\n<h3 class=\"wp-block-heading has-medium-grey-color has-text-color\"><strong>Diagram showing the nesting of artificial intelligence concepts.<\/strong><\/h3>\n\n\n\n<figure class=\"wp-block-image aligncenter size-large is-resized\"><img loading=\"lazy\" decoding=\"async\" width=\"1022\" height=\"1024\" src=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_03_schema_IA-1022x1024.png\" alt=\"\" class=\"wp-image-739\" style=\"width:378px;height:379px\" srcset=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_03_schema_IA-1022x1024.png 1022w, https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_03_schema_IA-300x300.png 300w, https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_03_schema_IA-150x150.png 150w, https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_03_schema_IA-768x769.png 768w, https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_03_schema_IA-1920x1924.png 1920w, https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_03_schema_IA-60x60.png 60w\" sizes=\"auto, (max-width: 1022px) 100vw, 1022px\" \/><figcaption class=\"wp-element-caption\"><strong><strong>Laetitia Mussard, inspired by \u201cCarto IA deepLearning\u201d, \u00a9 Bouliech, 2018, CC-BY-SA (source: Wikimedia Commons).<\/strong><\/strong><\/figcaption><\/figure>\n\n\n\n<p class=\"has-medium-grey-color has-text-color\"><strong>Text adapted from an article published in <a rel=\"noreferrer noopener\" href=\"https:\/\/ecoledesponts.fr\/sites\/ecoledesponts.fr\/files\/documents\/cdp_num1_ia.pdf\" target=\"_blank\">Le Cahier des Ponts n\u00b01<\/a>, L&#8217;intelligence artificielle, March 2019<\/strong>.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Before we can speak of cybernetics, Artificial Intelligence (AI) requires human intelligence, and that is what drives the research teams [&hellip;]<\/p>\n","protected":false},"author":6,"featured_media":1464,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_related_content_post":[],"_related_content_subject":[690],"_related_content_author":[946],"_related_content_category":[1716],"_related_content_folder":[],"_excerpt":"Artificial intelligence (AI) has been around for decades, but recently it has experienced significant advances in a growing number of industries. From the placement of advertising spots on a television channel to airline company schedules and even the optimal management of hydroelectric dam valleys, AI research has a bright future ahead of it.","_duration":6,"_manual_duration":false,"footnotes":""},"article-types":[13],"class_list":["post-1769","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","article-types-article"],"has_blocks":true,"block_data":[{"blockName":"enpc\/excerpt","attrs":{"lock":[],"metadata":[],"className":"","style":""},"innerBlocks":[],"innerHTML":"","innerContent":[],"rendered":""},{"blockName":"core\/image","attrs":{"id":863,"sizeSlug":"large","linkDestination":"none","align":"wide","className":"is-style-default","blob":"","url":"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_01_Grande_illu-1024x589.jpg","alt":"","caption":"\"Artificial Intelligence, Thinking Machines and the Future of Humanity\" \u00a9 Gerd Leonhard, CC-BY-SA (source : flickr)","lightbox":[],"title":"","href":"","rel":"","linkClass":"","width":"","height":"","aspectRatio":"","scale":"","linkTarget":"","lock":[],"metadata":[],"style":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<figure class=\"wp-block-image alignwide size-large is-style-default\"><img src=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_01_Grande_illu-1024x589.jpg\" alt=\"\" class=\"wp-image-863\"\/><figcaption class=\"wp-element-caption\">\"Artificial Intelligence, Thinking Machines and the Future of Humanity\" \u00a9 Gerd Leonhard, CC-BY-SA (source : flickr)<\/figcaption><\/figure>\n","innerContent":["\n<figure class=\"wp-block-image alignwide size-large is-style-default\"><img src=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_01_Grande_illu-1024x589.jpg\" alt=\"\" class=\"wp-image-863\"\/><figcaption class=\"wp-element-caption\">\"Artificial Intelligence, Thinking Machines and the Future of Humanity\" \u00a9 Gerd Leonhard, CC-BY-SA (source : flickr)<\/figcaption><\/figure>\n"],"rendered":"\n<figure class=\"wp-block-image alignwide size-large is-style-default\"><img src=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_01_Grande_illu-1024x589.jpg\" alt=\"\" class=\"wp-image-863\"\/><figcaption class=\"wp-element-caption\">\"Artificial Intelligence, Thinking Machines and the Future of Humanity\" \u00a9 Gerd Leonhard, CC-BY-SA (source : flickr)<\/figcaption><\/figure>\n"},{"blockName":"core\/paragraph","attrs":{"align":"","content":"Before we can speak of cybernetics, Artificial Intelligence (AI) requires human intelligence, and that is what drives the research teams at \u00c9cole nationale des ponts et chauss\u00e9es. Like his peers who are working to push the limits of AI in France and around the world, Axel Parmentier, researcher at CERMICS (Center of Teaching and Research in Mathematics and Scientific Calculation), is one of the players in this scientific and technological revolution that is shaking up countless activities: from the digital economy to the autonomous vehicle, including the optimization of industrial processes. \u201cWe have inherited more than a century of tradition and very strong skills in applied mathematics and at the same time, we are the actors of a scientific and technological discipline that is buzzing,\" explains the young researcher.","dropCap":false,"placeholder":"","direction":"","lock":[],"metadata":[],"className":"","style":"","backgroundColor":"","textColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<p>Before we can speak of cybernetics, Artificial Intelligence (AI) requires human intelligence, and that is what drives the research teams at \u00c9cole nationale des ponts et chauss\u00e9es. Like his peers who are working to push the limits of AI in France and around the world, Axel Parmentier, researcher at <a href=\"https:\/\/cermics-lab.enpc.fr\/\" target=\"_blank\" rel=\"noreferrer noopener\">CERMICS <\/a>(Center of Teaching and Research in Mathematics and Scientific Calculation), is one of the players in this scientific and technological revolution that is shaking up countless activities: from the digital economy to the autonomous vehicle, including the optimization of industrial processes. \u201cWe have inherited more than a century of tradition and very strong skills in applied mathematics and at the same time, we are the actors of a scientific and technological discipline that is buzzing,\" explains the young researcher.<\/p>\n","innerContent":["\n<p>Before we can speak of cybernetics, Artificial Intelligence (AI) requires human intelligence, and that is what drives the research teams at \u00c9cole nationale des ponts et chauss\u00e9es. Like his peers who are working to push the limits of AI in France and around the world, Axel Parmentier, researcher at <a href=\"https:\/\/cermics-lab.enpc.fr\/\" target=\"_blank\" rel=\"noreferrer noopener\">CERMICS <\/a>(Center of Teaching and Research in Mathematics and Scientific Calculation), is one of the players in this scientific and technological revolution that is shaking up countless activities: from the digital economy to the autonomous vehicle, including the optimization of industrial processes. \u201cWe have inherited more than a century of tradition and very strong skills in applied mathematics and at the same time, we are the actors of a scientific and technological discipline that is buzzing,\" explains the young researcher.<\/p>\n"],"rendered":"\n<p>Before we can speak of cybernetics, Artificial Intelligence (AI) requires human intelligence, and that is what drives the research teams at \u00c9cole nationale des ponts et chauss\u00e9es. Like his peers who are working to push the limits of AI in France and around the world, Axel Parmentier, researcher at <a href=\"https:\/\/cermics-lab.enpc.fr\/\" target=\"_blank\" rel=\"noreferrer noopener\">CERMICS <\/a>(Center of Teaching and Research in Mathematics and Scientific Calculation), is one of the players in this scientific and technological revolution that is shaking up countless activities: from the digital economy to the autonomous vehicle, including the optimization of industrial processes. \u201cWe have inherited more than a century of tradition and very strong skills in applied mathematics and at the same time, we are the actors of a scientific and technological discipline that is buzzing,\" explains the young researcher.<\/p>\n"},{"blockName":"core\/paragraph","attrs":{"align":"","content":"Artificial intelligence today covers three essential types of applications:","dropCap":false,"placeholder":"","direction":"","lock":[],"metadata":[],"className":"","style":"","backgroundColor":"","textColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<p>Artificial intelligence today covers three essential types of applications:<\/p>\n","innerContent":["\n<p>Artificial intelligence today covers three essential types of applications:<\/p>\n"],"rendered":"\n<p>Artificial intelligence today covers three essential types of applications:<\/p>\n"},{"blockName":"core\/list","attrs":{"ordered":false,"values":"","type":"","start":0,"reversed":false,"placeholder":"","lock":[],"metadata":[],"className":"wp-block-list","style":"","backgroundColor":"","textColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[{"blockName":"core\/list-item","attrs":{"placeholder":"","content":"\u00a0The first is data-mining, describing gigantic sets of data in order to extract a relevant summary. This applies to the bulk of applications that are now mature in the traditional economy,\u201d says Axel Parmentier ;","lock":[],"metadata":[],"className":"","style":"","backgroundColor":"","textColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<li>&nbsp;The first is data-mining, describing gigantic sets of data in order to extract a relevant summary. This applies to the bulk of applications that are now mature in the traditional economy,\u201d says Axel Parmentier ;<\/li>\n","innerContent":["\n<li>&nbsp;The first is data-mining, describing gigantic sets of data in order to extract a relevant summary. This applies to the bulk of applications that are now mature in the traditional economy,\u201d says Axel Parmentier ;<\/li>\n"],"rendered":"\n<li>&nbsp;The first is data-mining, describing gigantic sets of data in order to extract a relevant summary. This applies to the bulk of applications that are now mature in the traditional economy,\u201d says Axel Parmentier ;<\/li>\n"}],"innerHTML":"\n<ul class=\"wp-block-list\"><\/ul>\n","innerContent":["\n<ul class=\"wp-block-list\">",null,"<\/ul>\n"],"rendered":"\n<ul class=\"wp-block-list\">\n<li>&nbsp;The first is data-mining, describing gigantic sets of data in order to extract a relevant summary. This applies to the bulk of applications that are now mature in the traditional economy,\u201d says Axel Parmentier ;<\/li>\n<\/ul>\n"},{"blockName":"core\/paragraph","attrs":{"align":"","content":"","dropCap":false,"placeholder":"","direction":"","lock":[],"metadata":[],"className":"","style":"","backgroundColor":"","textColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<p><\/p>\n","innerContent":["\n<p><\/p>\n"],"rendered":"\n<p><\/p>\n"},{"blockName":"core\/list","attrs":{"ordered":false,"values":"","type":"","start":0,"reversed":false,"placeholder":"","lock":[],"metadata":[],"className":"wp-block-list","style":"","backgroundColor":"","textColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[{"blockName":"core\/list-item","attrs":{"placeholder":"","content":"The second is the use of data to make predictions through supervised learning. It is notably a subject of expertise of the IMAGINE team (at the Gaspard-Monge computer laboratory), particularly in the field of computer vision ;","lock":[],"metadata":[],"className":"","style":"","backgroundColor":"","textColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<li>The second is the use of data to make predictions through supervised learning. It is notably a subject of expertise of the IMAGINE team (at the Gaspard-Monge computer laboratory), particularly in the field of computer vision ;<\/li>\n","innerContent":["\n<li>The second is the use of data to make predictions through supervised learning. It is notably a subject of expertise of the IMAGINE team (at the Gaspard-Monge computer laboratory), particularly in the field of computer vision ;<\/li>\n"],"rendered":"\n<li>The second is the use of data to make predictions through supervised learning. It is notably a subject of expertise of the IMAGINE team (at the Gaspard-Monge computer laboratory), particularly in the field of computer vision ;<\/li>\n"}],"innerHTML":"\n<ul class=\"wp-block-list\"><\/ul>\n","innerContent":["\n<ul class=\"wp-block-list\">",null,"<\/ul>\n"],"rendered":"\n<ul class=\"wp-block-list\">\n<li>The second is the use of data to make predictions through supervised learning. It is notably a subject of expertise of the IMAGINE team (at the Gaspard-Monge computer laboratory), particularly in the field of computer vision ;<\/li>\n<\/ul>\n"},{"blockName":"core\/paragraph","attrs":{"align":"","content":"","dropCap":false,"placeholder":"","direction":"","lock":[],"metadata":[],"className":"","style":"","backgroundColor":"","textColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<p><\/p>\n","innerContent":["\n<p><\/p>\n"],"rendered":"\n<p><\/p>\n"},{"blockName":"core\/list","attrs":{"ordered":false,"values":"","type":"","start":0,"reversed":false,"placeholder":"","lock":[],"metadata":[],"className":"wp-block-list","style":"","backgroundColor":"","textColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[{"blockName":"core\/list-item","attrs":{"placeholder":"","content":"The third is decision-making support, which will make it possible to recommend the best possible decision on complex problems using data-driven optimization models, and on which the CERMICS optimization team is working in particular.","lock":[],"metadata":[],"className":"","style":"","backgroundColor":"","textColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<li>The third is decision-making support, which will make it possible to recommend the best possible decision on complex problems using data-driven optimization models, and on which the CERMICS optimization team is working in particular.<\/li>\n","innerContent":["\n<li>The third is decision-making support, which will make it possible to recommend the best possible decision on complex problems using data-driven optimization models, and on which the CERMICS optimization team is working in particular.<\/li>\n"],"rendered":"\n<li>The third is decision-making support, which will make it possible to recommend the best possible decision on complex problems using data-driven optimization models, and on which the CERMICS optimization team is working in particular.<\/li>\n"}],"innerHTML":"\n<ul class=\"wp-block-list\"><\/ul>\n","innerContent":["\n<ul class=\"wp-block-list\">",null,"<\/ul>\n"],"rendered":"\n<ul class=\"wp-block-list\">\n<li>The third is decision-making support, which will make it possible to recommend the best possible decision on complex problems using data-driven optimization models, and on which the CERMICS optimization team is working in particular.<\/li>\n<\/ul>\n"},{"blockName":"core\/heading","attrs":{"textColor":"red","textAlign":"","content":"Operational research, a formidable decision support tool","level":2,"levelOptions":[],"placeholder":"","lock":[],"metadata":[],"align":"","className":"wp-block-heading has-red-color has-text-color","style":"","backgroundColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>Operational research, a formidable decision support tool<\/strong><\/h2>\n","innerContent":["\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>Operational research, a formidable decision support tool<\/strong><\/h2>\n"],"rendered":"\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>Operational research, a formidable decision support tool<\/strong><\/h2>\n"},{"blockName":"core\/paragraph","attrs":{"align":"","content":"In this large AI family, Axel Parmentier, is an expert in operational research. This discipline combines applied mathematics and computer science to help industrialists allocate their resources more effectively. For example, it provides tools for planning a supply chain or managing energy networks (smart grids).","dropCap":false,"placeholder":"","direction":"","lock":[],"metadata":[],"className":"","style":"","backgroundColor":"","textColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<p>In this large AI family, Axel Parmentier, is an expert in operational research. This discipline combines applied mathematics and computer science to help industrialists allocate their resources more effectively. For example, it provides tools for planning a supply chain or managing energy networks (smart grids).<\/p>\n","innerContent":["\n<p>In this large AI family, Axel Parmentier, is an expert in operational research. This discipline combines applied mathematics and computer science to help industrialists allocate their resources more effectively. For example, it provides tools for planning a supply chain or managing energy networks (smart grids).<\/p>\n"],"rendered":"\n<p>In this large AI family, Axel Parmentier, is an expert in operational research. This discipline combines applied mathematics and computer science to help industrialists allocate their resources more effectively. For example, it provides tools for planning a supply chain or managing energy networks (smart grids).<\/p>\n"},{"blockName":"core\/paragraph","attrs":{"align":"","content":"It is implemented within the school through a chair with Air France inaugurated at the end of 2016, entitled \u201cArtificial intelligence for the Air Transport industry\u201d, the first in France to focus on the interface between these disciplines. One of the creations by the research unit, along with the airline's R&D teams, is an algorithm that builds flight sequences for aircraft and crews and enables these flights to be operated at lower cost. This is a tricky issue for commercial software.On this type of problem, it is not uncommon for an algorithm to be able to solve problems for 500 flights in a few seconds, problems for 600 flights in a few hours, but be totally unable to solve problems for 700 flights. \u201cThis barrier is explained not by reasons of computing power, but by a wall of theoretical complexity\u201d explains Axel Parmentier. Therefore, mathematical models and algorithms have to be defined to find solutions, which makes great thesis topics for our laboratory!\u201d.\u00a0","dropCap":false,"placeholder":"","direction":"","lock":[],"metadata":[],"className":"","style":"","backgroundColor":"","textColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<p>It is implemented within the school through a chair with Air France inaugurated at the end of 2016, entitled \u201c<a href=\"https:\/\/ecoledesponts.fr\/intelligence-artificielle-pour-laerien\" data-type=\"URL\" data-id=\"https:\/\/ecoledesponts.fr\/intelligence-artificielle-pour-laerien\" target=\"_blank\" rel=\"noreferrer noopener\">Artificial intelligence for the Air Transport industry<\/a>\u201d, the first in France to focus on the interface between these disciplines. One of the creations by the research unit, along with the airline's R&amp;D teams, is an algorithm that builds flight sequences for aircraft and crews and enables these flights to be operated at lower cost. This is a tricky issue for commercial software.On this type of problem, it is not uncommon for an algorithm to be able to solve problems for 500 flights in a few seconds, problems for 600 flights in a few hours, but be totally unable to solve problems for 700 flights<em>. <\/em>\u201cThis barrier is explained not by reasons of computing power, but by a wall of theoretical complexity\u201d explains Axel Parmentier. Therefore, mathematical models and algorithms have to be defined to find solutions, which makes great thesis topics for our laboratory!\u201d.&nbsp;<\/p>\n","innerContent":["\n<p>It is implemented within the school through a chair with Air France inaugurated at the end of 2016, entitled \u201c<a href=\"https:\/\/ecoledesponts.fr\/intelligence-artificielle-pour-laerien\" data-type=\"URL\" data-id=\"https:\/\/ecoledesponts.fr\/intelligence-artificielle-pour-laerien\" target=\"_blank\" rel=\"noreferrer noopener\">Artificial intelligence for the Air Transport industry<\/a>\u201d, the first in France to focus on the interface between these disciplines. One of the creations by the research unit, along with the airline's R&amp;D teams, is an algorithm that builds flight sequences for aircraft and crews and enables these flights to be operated at lower cost. This is a tricky issue for commercial software.On this type of problem, it is not uncommon for an algorithm to be able to solve problems for 500 flights in a few seconds, problems for 600 flights in a few hours, but be totally unable to solve problems for 700 flights<em>. <\/em>\u201cThis barrier is explained not by reasons of computing power, but by a wall of theoretical complexity\u201d explains Axel Parmentier. Therefore, mathematical models and algorithms have to be defined to find solutions, which makes great thesis topics for our laboratory!\u201d.&nbsp;<\/p>\n"],"rendered":"\n<p>It is implemented within the school through a chair with Air France inaugurated at the end of 2016, entitled \u201c<a href=\"https:\/\/ecoledesponts.fr\/intelligence-artificielle-pour-laerien\" data-type=\"URL\" data-id=\"https:\/\/ecoledesponts.fr\/intelligence-artificielle-pour-laerien\" target=\"_blank\" rel=\"noreferrer noopener\">Artificial intelligence for the Air Transport industry<\/a>\u201d, the first in France to focus on the interface between these disciplines. One of the creations by the research unit, along with the airline's R&amp;D teams, is an algorithm that builds flight sequences for aircraft and crews and enables these flights to be operated at lower cost. This is a tricky issue for commercial software.On this type of problem, it is not uncommon for an algorithm to be able to solve problems for 500 flights in a few seconds, problems for 600 flights in a few hours, but be totally unable to solve problems for 700 flights<em>. <\/em>\u201cThis barrier is explained not by reasons of computing power, but by a wall of theoretical complexity\u201d explains Axel Parmentier. Therefore, mathematical models and algorithms have to be defined to find solutions, which makes great thesis topics for our laboratory!\u201d.&nbsp;<\/p>\n"},{"blockName":"core\/heading","attrs":{"level":3,"textColor":"medium-grey","textAlign":"","content":"Frank-Wolfe Algorithm","levelOptions":[],"placeholder":"","lock":[],"metadata":[],"align":"","className":"wp-block-heading has-medium-grey-color has-text-color","style":"","backgroundColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<h3 class=\"wp-block-heading has-medium-grey-color has-text-color\"><strong>Frank-Wolfe Algorithm<\/strong><\/h3>\n","innerContent":["\n<h3 class=\"wp-block-heading has-medium-grey-color has-text-color\"><strong>Frank-Wolfe Algorithm<\/strong><\/h3>\n"],"rendered":"\n<h3 class=\"wp-block-heading has-medium-grey-color has-text-color\"><strong>Frank-Wolfe Algorithm<\/strong><\/h3>\n"},{"blockName":"core\/image","attrs":{"id":737,"width":"414px","height":"365px","sizeSlug":"large","linkDestination":"none","align":"center","blob":"","url":"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_02_Franck-Wolfe-1024x904.jpg","alt":"","caption":"In operational research, the aim is to find the best decision; in machine learning, it is a matter of finding the model that makes the best predictions. Mathematically, this amounts to finding the minimum of a function, here the point at the bottom of the blue sphere. Many efficient algorithms, such as the Franck-Wolfe algorithm illustrated here, are based on geometry. \"Frank-Wolfe Algorithm\" \u00a9 Stephanie Stutz, CC-BY-SA (source: Wikimedia Commons).","lightbox":[],"title":"","href":"","rel":"","linkClass":"","aspectRatio":"","scale":"","linkTarget":"","lock":[],"metadata":[],"className":"wp-block-image aligncenter size-large is-resized","style":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<figure class=\"wp-block-image aligncenter size-large is-resized\"><img src=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_02_Franck-Wolfe-1024x904.jpg\" alt=\"\" class=\"wp-image-737\" style=\"width:414px;height:365px\"\/><figcaption class=\"wp-element-caption\"><strong>In operational research, the aim is to find the best decision; in machine learning, it is a matter of finding the model that makes the best predictions. Mathematically, this amounts to finding the minimum of a function, here the point at the bottom of the blue sphere. Many efficient algorithms, such as the Franck-Wolfe algorithm illustrated here, are based on geometry. \"Frank-Wolfe Algorithm\" \u00a9 Stephanie Stutz, CC-BY-SA (source: Wikimedia Commons).<\/strong><\/figcaption><\/figure>\n","innerContent":["\n<figure class=\"wp-block-image aligncenter size-large is-resized\"><img src=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_02_Franck-Wolfe-1024x904.jpg\" alt=\"\" class=\"wp-image-737\" style=\"width:414px;height:365px\"\/><figcaption class=\"wp-element-caption\"><strong>In operational research, the aim is to find the best decision; in machine learning, it is a matter of finding the model that makes the best predictions. Mathematically, this amounts to finding the minimum of a function, here the point at the bottom of the blue sphere. Many efficient algorithms, such as the Franck-Wolfe algorithm illustrated here, are based on geometry. \"Frank-Wolfe Algorithm\" \u00a9 Stephanie Stutz, CC-BY-SA (source: Wikimedia Commons).<\/strong><\/figcaption><\/figure>\n"],"rendered":"\n<figure class=\"wp-block-image aligncenter size-large is-resized\"><img src=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_02_Franck-Wolfe-1024x904.jpg\" alt=\"\" class=\"wp-image-737\" style=\"width:414px;height:365px\"\/><figcaption class=\"wp-element-caption\"><strong>In operational research, the aim is to find the best decision; in machine learning, it is a matter of finding the model that makes the best predictions. Mathematically, this amounts to finding the minimum of a function, here the point at the bottom of the blue sphere. Many efficient algorithms, such as the Franck-Wolfe algorithm illustrated here, are based on geometry. \"Frank-Wolfe Algorithm\" \u00a9 Stephanie Stutz, CC-BY-SA (source: Wikimedia Commons).<\/strong><\/figcaption><\/figure>\n"},{"blockName":"core\/heading","attrs":{"textColor":"red","textAlign":"","content":"Significant progress thanks to increasingly efficient algorithms","level":2,"levelOptions":[],"placeholder":"","lock":[],"metadata":[],"align":"","className":"wp-block-heading has-red-color has-text-color","style":"","backgroundColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>Significant progress thanks to increasingly efficient algorithms<\/strong><\/h2>\n","innerContent":["\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>Significant progress thanks to increasingly efficient algorithms<\/strong><\/h2>\n"],"rendered":"\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>Significant progress thanks to increasingly efficient algorithms<\/strong><\/h2>\n"},{"blockName":"core\/paragraph","attrs":{"align":"","content":"AI has been around for decades, but it has recently seen significant advances in a growing number of sectors. Why? \u201cThese advances are made possible by the increase in computer power or the availability of big data, but the key lies in the development of increasingly high-performance algorithms, i.e. in the work of researchers such as those in our laboratory\u201d says Axel Parmentier.","dropCap":false,"placeholder":"","direction":"","lock":[],"metadata":[],"className":"","style":"","backgroundColor":"","textColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<p>AI has been around for decades, but it has recently seen significant advances in a growing number of sectors. Why? \u201cThese advances are made possible by the increase in computer power or the availability of big data, but the key lies in the development of increasingly high-performance algorithms, i.e. in the work of researchers such as those in our laboratory\u201d says Axel Parmentier.<\/p>\n","innerContent":["\n<p>AI has been around for decades, but it has recently seen significant advances in a growing number of sectors. Why? \u201cThese advances are made possible by the increase in computer power or the availability of big data, but the key lies in the development of increasingly high-performance algorithms, i.e. in the work of researchers such as those in our laboratory\u201d says Axel Parmentier.<\/p>\n"],"rendered":"\n<p>AI has been around for decades, but it has recently seen significant advances in a growing number of sectors. Why? \u201cThese advances are made possible by the increase in computer power or the availability of big data, but the key lies in the development of increasingly high-performance algorithms, i.e. in the work of researchers such as those in our laboratory\u201d says Axel Parmentier.<\/p>\n"},{"blockName":"core\/paragraph","attrs":{"align":"","content":"The algorithm is, in his words, \u201cthe computer equivalent of a pastry chef's recipe\u201d which, through a succession of elementary mathematical operations, allows the computer to perform tasks such as making decisions or predictions. A line of code in software is a basic operation. This is what researchers do on a daily basis! Take the example of the AlphaGo Zero algorithm by the company DeepMind which beat the human brain in 2015 in the game of Go, which is no more than a sequence of predictions and decisions.","dropCap":false,"placeholder":"","direction":"","lock":[],"metadata":[],"className":"","style":"","backgroundColor":"","textColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<p>The algorithm is, in his words, \u201cthe computer equivalent of a pastry chef's recipe\u201d which, through a succession of elementary mathematical operations, allows the computer to perform tasks such as making decisions or predictions. A line of code in software is a basic operation. This is what researchers do on a daily basis! Take the example of the AlphaGo Zero algorithm by the company DeepMind which beat the human brain in 2015 in the game of Go, which is no more than a sequence of predictions and decisions.<\/p>\n","innerContent":["\n<p>The algorithm is, in his words, \u201cthe computer equivalent of a pastry chef's recipe\u201d which, through a succession of elementary mathematical operations, allows the computer to perform tasks such as making decisions or predictions. A line of code in software is a basic operation. This is what researchers do on a daily basis! Take the example of the AlphaGo Zero algorithm by the company DeepMind which beat the human brain in 2015 in the game of Go, which is no more than a sequence of predictions and decisions.<\/p>\n"],"rendered":"\n<p>The algorithm is, in his words, \u201cthe computer equivalent of a pastry chef's recipe\u201d which, through a succession of elementary mathematical operations, allows the computer to perform tasks such as making decisions or predictions. A line of code in software is a basic operation. This is what researchers do on a daily basis! Take the example of the AlphaGo Zero algorithm by the company DeepMind which beat the human brain in 2015 in the game of Go, which is no more than a sequence of predictions and decisions.<\/p>\n"},{"blockName":"core\/paragraph","attrs":{"align":"","content":"\u201cAn enormous amount of progress has been made in algorithms in recent decades. Take the example of deep learning: 20 years ago, the models didn\u2019t work very well, partly because there was a lack of data to exploit, but mainly because proper neural network architectures were not yet available. For example, running the 1990 integer linear programming algorithms (a flagship tool in operations research) with today's computers can increase speed by hundreds of thousands of times over the computers of the day. With the latest generation of algorithms, the progress is by thousands of billions of times! Advances in learning and computer vision algorithms have led to advances in automatic recognition of objects in video sequences, a technology used in the autonomous car.","dropCap":false,"placeholder":"","direction":"","lock":[],"metadata":[],"className":"","style":"","backgroundColor":"","textColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<p>\u201cAn enormous amount of progress has been made in algorithms in recent decades. Take the example of deep learning: 20 years ago, the models didn\u2019t work very well, partly because there was a lack of data to exploit, but mainly because proper neural network architectures were not yet available. For example, running the 1990 integer linear programming algorithms (a flagship tool in operations research) with today's computers can increase speed by hundreds of thousands of times over the computers of the day. With the latest generation of algorithms, the progress is by thousands of billions of times! Advances in learning and computer vision algorithms have led to advances in automatic recognition of objects in video sequences, a technology used in the autonomous car.<\/p>\n","innerContent":["\n<p>\u201cAn enormous amount of progress has been made in algorithms in recent decades. Take the example of deep learning: 20 years ago, the models didn\u2019t work very well, partly because there was a lack of data to exploit, but mainly because proper neural network architectures were not yet available. For example, running the 1990 integer linear programming algorithms (a flagship tool in operations research) with today's computers can increase speed by hundreds of thousands of times over the computers of the day. With the latest generation of algorithms, the progress is by thousands of billions of times! Advances in learning and computer vision algorithms have led to advances in automatic recognition of objects in video sequences, a technology used in the autonomous car.<\/p>\n"],"rendered":"\n<p>\u201cAn enormous amount of progress has been made in algorithms in recent decades. Take the example of deep learning: 20 years ago, the models didn\u2019t work very well, partly because there was a lack of data to exploit, but mainly because proper neural network architectures were not yet available. For example, running the 1990 integer linear programming algorithms (a flagship tool in operations research) with today's computers can increase speed by hundreds of thousands of times over the computers of the day. With the latest generation of algorithms, the progress is by thousands of billions of times! Advances in learning and computer vision algorithms have led to advances in automatic recognition of objects in video sequences, a technology used in the autonomous car.<\/p>\n"},{"blockName":"core\/heading","attrs":{"textColor":"red","textAlign":"","content":"A multitude of unsuspected applications","level":2,"levelOptions":[],"placeholder":"","lock":[],"metadata":[],"align":"","className":"wp-block-heading has-red-color has-text-color","style":"","backgroundColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>A multitude of unsuspected applications<\/strong><\/h2>\n","innerContent":["\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>A multitude of unsuspected applications<\/strong><\/h2>\n"],"rendered":"\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>A multitude of unsuspected applications<\/strong><\/h2>\n"},{"blockName":"core\/paragraph","attrs":{"align":"","content":"\u00c9cole nationale des ponts et chauss\u00e9es contributes to the progress of research by developing new models and improving the understanding of those produced by the scientific community. Its industrial partners, such as Air France, Renault and Total, are looking for algorithms that can solve their problems, but above all the scientific guarantee of the result. \u201cBefore validating algorithms that enable the running of a power plant or the movement of a car, we must be certain of the quality of the results they produce, which requires an understanding of their mathematical properties,\u201d explains the researcher. \u201cThis is the beauty of mathematics: you model a concrete problem in the form of an abstract mathematical problem, you build a nice theory to solve this problem and you obtain efficient algorithms that find a multitude of unsuspected applications\u201d. And to quote linear integer programming, which makes it possible to model a wide variety of problems. Today, it has applications ranging from the placement of advertising spots on a television channel to airline company schedules, including the optimal management of hydroelectric dam valleys, or the organization of production for an industrial company.","dropCap":false,"placeholder":"","direction":"","lock":[],"metadata":[],"className":"","style":"","backgroundColor":"","textColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<p>\u00c9cole nationale des ponts et chauss\u00e9es contributes to the progress of research by developing new models and improving the understanding of those produced by the scientific community. Its industrial partners, such as Air France, Renault and Total, are looking for algorithms that can solve their problems, but above all the scientific guarantee of the result. \u201cBefore validating algorithms that enable the running of a power plant or the movement of a car, we must be certain of the quality of the results they produce, which requires an understanding of their mathematical properties,\u201d explains the researcher. \u201cThis is the beauty of mathematics: you model a concrete problem in the form of an abstract mathematical problem, you build a nice theory to solve this problem and you obtain efficient algorithms that find a multitude of unsuspected applications\u201d. And to quote linear integer programming, which makes it possible to model a wide variety of problems. Today, it has applications ranging from the placement of advertising spots on a television channel to airline company schedules, including the optimal management of hydroelectric dam valleys, or the organization of production for an industrial company.<\/p>\n","innerContent":["\n<p>\u00c9cole nationale des ponts et chauss\u00e9es contributes to the progress of research by developing new models and improving the understanding of those produced by the scientific community. Its industrial partners, such as Air France, Renault and Total, are looking for algorithms that can solve their problems, but above all the scientific guarantee of the result. \u201cBefore validating algorithms that enable the running of a power plant or the movement of a car, we must be certain of the quality of the results they produce, which requires an understanding of their mathematical properties,\u201d explains the researcher. \u201cThis is the beauty of mathematics: you model a concrete problem in the form of an abstract mathematical problem, you build a nice theory to solve this problem and you obtain efficient algorithms that find a multitude of unsuspected applications\u201d. And to quote linear integer programming, which makes it possible to model a wide variety of problems. Today, it has applications ranging from the placement of advertising spots on a television channel to airline company schedules, including the optimal management of hydroelectric dam valleys, or the organization of production for an industrial company.<\/p>\n"],"rendered":"\n<p>\u00c9cole nationale des ponts et chauss\u00e9es contributes to the progress of research by developing new models and improving the understanding of those produced by the scientific community. Its industrial partners, such as Air France, Renault and Total, are looking for algorithms that can solve their problems, but above all the scientific guarantee of the result. \u201cBefore validating algorithms that enable the running of a power plant or the movement of a car, we must be certain of the quality of the results they produce, which requires an understanding of their mathematical properties,\u201d explains the researcher. \u201cThis is the beauty of mathematics: you model a concrete problem in the form of an abstract mathematical problem, you build a nice theory to solve this problem and you obtain efficient algorithms that find a multitude of unsuspected applications\u201d. And to quote linear integer programming, which makes it possible to model a wide variety of problems. Today, it has applications ranging from the placement of advertising spots on a television channel to airline company schedules, including the optimal management of hydroelectric dam valleys, or the organization of production for an industrial company.<\/p>\n"},{"blockName":"core\/heading","attrs":{"textColor":"red","textAlign":"","content":"Cross-research in the heart of the laboratories","level":2,"levelOptions":[],"placeholder":"","lock":[],"metadata":[],"align":"","className":"wp-block-heading has-red-color has-text-color","style":"","backgroundColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>Cross-research in the heart of the laboratories<\/strong><strong><\/strong><\/h2>\n","innerContent":["\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>Cross-research in the heart of the laboratories<\/strong><strong><\/strong><\/h2>\n"],"rendered":"\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>Cross-research in the heart of the laboratories<\/strong><strong><\/strong><\/h2>\n"},{"blockName":"core\/paragraph","attrs":{"align":"","content":"The needs of the economy and scientific progress even lead to the mixing of disciplines within its laboratories. This is the case, for example, between CERMICS and IMAGINE, which traditionally operate on distinct fields: probability, numerical analysis and optimization for the former, statistics, prediction, artificial vision and machine learning for the latter. The evolution of technology, as in the reinforcement learning popularized by AlphaGo, blurs the boundaries between optimization and machine learning. \u201cThis encourages crossed research between our teams. This work is applied by our industrial partner Air France to predictive aircraft maintenance, i.e. the ability to use the data fed back by the aircraft to anticipate accurately when a particular part of an aircraft is likely to malfunction and thus avoid breakdowns\u201d, notes Axel Parmentier.By combining the fields of prediction and decision making, their common goal is to advance practical and theoretical knowledge.","dropCap":false,"placeholder":"","direction":"","lock":[],"metadata":[],"className":"","style":"","backgroundColor":"","textColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<p>The needs of the economy and scientific progress even lead to the mixing of disciplines within its laboratories. This is the case, for example, between CERMICS and IMAGINE, which traditionally operate on distinct fields: probability, numerical analysis and optimization for the former, statistics, prediction, artificial vision and machine learning for the latter. The evolution of technology, as in the reinforcement learning popularized by AlphaGo, blurs the boundaries between optimization and machine learning. \u201cThis encourages crossed research between our teams. This work is applied by our industrial partner Air France to predictive aircraft maintenance, i.e. the ability to use the data fed back by the aircraft to anticipate accurately when a particular part of an aircraft is likely to malfunction and thus avoid breakdowns\u201d, notes Axel Parmentier.By combining the fields of prediction and decision making, their common goal is to advance practical and theoretical knowledge.<\/p>\n","innerContent":["\n<p>The needs of the economy and scientific progress even lead to the mixing of disciplines within its laboratories. This is the case, for example, between CERMICS and IMAGINE, which traditionally operate on distinct fields: probability, numerical analysis and optimization for the former, statistics, prediction, artificial vision and machine learning for the latter. The evolution of technology, as in the reinforcement learning popularized by AlphaGo, blurs the boundaries between optimization and machine learning. \u201cThis encourages crossed research between our teams. This work is applied by our industrial partner Air France to predictive aircraft maintenance, i.e. the ability to use the data fed back by the aircraft to anticipate accurately when a particular part of an aircraft is likely to malfunction and thus avoid breakdowns\u201d, notes Axel Parmentier.By combining the fields of prediction and decision making, their common goal is to advance practical and theoretical knowledge.<\/p>\n"],"rendered":"\n<p>The needs of the economy and scientific progress even lead to the mixing of disciplines within its laboratories. This is the case, for example, between CERMICS and IMAGINE, which traditionally operate on distinct fields: probability, numerical analysis and optimization for the former, statistics, prediction, artificial vision and machine learning for the latter. The evolution of technology, as in the reinforcement learning popularized by AlphaGo, blurs the boundaries between optimization and machine learning. \u201cThis encourages crossed research between our teams. This work is applied by our industrial partner Air France to predictive aircraft maintenance, i.e. the ability to use the data fed back by the aircraft to anticipate accurately when a particular part of an aircraft is likely to malfunction and thus avoid breakdowns\u201d, notes Axel Parmentier.By combining the fields of prediction and decision making, their common goal is to advance practical and theoretical knowledge.<\/p>\n"},{"blockName":"core\/heading","attrs":{"textColor":"red","textAlign":"","content":"Towards a quest for excellence","level":2,"levelOptions":[],"placeholder":"","lock":[],"metadata":[],"align":"","className":"wp-block-heading has-red-color has-text-color","style":"","backgroundColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>Towards a quest for excellence<\/strong><\/h2>\n","innerContent":["\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>Towards a quest for excellence<\/strong><\/h2>\n"],"rendered":"\n<h2 class=\"wp-block-heading has-red-color has-text-color\"><strong>Towards a quest for excellence<\/strong><\/h2>\n"},{"blockName":"core\/paragraph","attrs":{"align":"","content":"\u201cIn the age of artificial intelligence, the potential applications of applied mathematics are almost infinite, but our top priority remains academic excellence,\u201d says Axel Parmentier. As in other fields of research, this quest for excellence is also measured by regular publications in top-ranking scientific journals, frequent participation in conferences that set the pace in this field, cooperation with European or American universities or close collaboration with institutes such as Inria (National Institute for Research in Digital Science and Technology). One of the CERMICS teams has thus developed first-rate expertise with this institute in the field of molecular simulation, which should become a major area of progress in the coming years for the chemical and pharmaceutical industries.","dropCap":false,"placeholder":"","direction":"","lock":[],"metadata":[],"className":"","style":"","backgroundColor":"","textColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<p>\u201cIn the age of artificial intelligence, the potential applications of applied mathematics are almost infinite, but our top priority remains academic excellence,\u201d says Axel Parmentier. As in other fields of research, this quest for excellence is also measured by regular publications in top-ranking scientific journals, frequent participation in conferences that set the pace in this field, cooperation with European or American universities or close collaboration with institutes such as <a href=\"https:\/\/www.inria.fr\/fr\" target=\"_blank\" rel=\"noreferrer noopener\">Inria<\/a> (National Institute for Research in Digital Science and Technology). One of the CERMICS teams has thus developed first-rate expertise with this institute in the field of molecular simulation, which should become a major area of progress in the coming years for the chemical and pharmaceutical industries.<\/p>\n","innerContent":["\n<p>\u201cIn the age of artificial intelligence, the potential applications of applied mathematics are almost infinite, but our top priority remains academic excellence,\u201d says Axel Parmentier. As in other fields of research, this quest for excellence is also measured by regular publications in top-ranking scientific journals, frequent participation in conferences that set the pace in this field, cooperation with European or American universities or close collaboration with institutes such as <a href=\"https:\/\/www.inria.fr\/fr\" target=\"_blank\" rel=\"noreferrer noopener\">Inria<\/a> (National Institute for Research in Digital Science and Technology). One of the CERMICS teams has thus developed first-rate expertise with this institute in the field of molecular simulation, which should become a major area of progress in the coming years for the chemical and pharmaceutical industries.<\/p>\n"],"rendered":"\n<p>\u201cIn the age of artificial intelligence, the potential applications of applied mathematics are almost infinite, but our top priority remains academic excellence,\u201d says Axel Parmentier. As in other fields of research, this quest for excellence is also measured by regular publications in top-ranking scientific journals, frequent participation in conferences that set the pace in this field, cooperation with European or American universities or close collaboration with institutes such as <a href=\"https:\/\/www.inria.fr\/fr\" target=\"_blank\" rel=\"noreferrer noopener\">Inria<\/a> (National Institute for Research in Digital Science and Technology). One of the CERMICS teams has thus developed first-rate expertise with this institute in the field of molecular simulation, which should become a major area of progress in the coming years for the chemical and pharmaceutical industries.<\/p>\n"},{"blockName":"core\/heading","attrs":{"level":3,"textColor":"medium-grey","textAlign":"","content":"Diagram showing the nesting of artificial intelligence concepts.","levelOptions":[],"placeholder":"","lock":[],"metadata":[],"align":"","className":"wp-block-heading has-medium-grey-color has-text-color","style":"","backgroundColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<h3 class=\"wp-block-heading has-medium-grey-color has-text-color\"><strong>Diagram showing the nesting of artificial intelligence concepts.<\/strong><\/h3>\n","innerContent":["\n<h3 class=\"wp-block-heading has-medium-grey-color has-text-color\"><strong>Diagram showing the nesting of artificial intelligence concepts.<\/strong><\/h3>\n"],"rendered":"\n<h3 class=\"wp-block-heading has-medium-grey-color has-text-color\"><strong>Diagram showing the nesting of artificial intelligence concepts.<\/strong><\/h3>\n"},{"blockName":"core\/image","attrs":{"id":739,"width":"378px","height":"379px","sizeSlug":"large","linkDestination":"none","align":"center","blob":"","url":"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_03_schema_IA-1022x1024.png","alt":"","caption":"Laetitia Mussard, inspired by \u201cCarto IA deepLearning\u201d, \u00a9 Bouliech, 2018, CC-BY-SA (source: Wikimedia Commons).","lightbox":[],"title":"","href":"","rel":"","linkClass":"","aspectRatio":"","scale":"","linkTarget":"","lock":[],"metadata":[],"className":"wp-block-image aligncenter size-large is-resized","style":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<figure class=\"wp-block-image aligncenter size-large is-resized\"><img src=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_03_schema_IA-1022x1024.png\" alt=\"\" class=\"wp-image-739\" style=\"width:378px;height:379px\"\/><figcaption class=\"wp-element-caption\"><strong><strong>Laetitia Mussard, inspired by \u201cCarto IA deepLearning\u201d, \u00a9 Bouliech, 2018, CC-BY-SA (source: Wikimedia Commons).<\/strong><\/strong><\/figcaption><\/figure>\n","innerContent":["\n<figure class=\"wp-block-image aligncenter size-large is-resized\"><img src=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_03_schema_IA-1022x1024.png\" alt=\"\" class=\"wp-image-739\" style=\"width:378px;height:379px\"\/><figcaption class=\"wp-element-caption\"><strong><strong>Laetitia Mussard, inspired by \u201cCarto IA deepLearning\u201d, \u00a9 Bouliech, 2018, CC-BY-SA (source: Wikimedia Commons).<\/strong><\/strong><\/figcaption><\/figure>\n"],"rendered":"\n<figure class=\"wp-block-image aligncenter size-large is-resized\"><img src=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2021\/12\/Cdp_Numero_1_02_Carte_blanche_Fig_03_schema_IA-1022x1024.png\" alt=\"\" class=\"wp-image-739\" style=\"width:378px;height:379px\"\/><figcaption class=\"wp-element-caption\"><strong><strong>Laetitia Mussard, inspired by \u201cCarto IA deepLearning\u201d, \u00a9 Bouliech, 2018, CC-BY-SA (source: Wikimedia Commons).<\/strong><\/strong><\/figcaption><\/figure>\n"},{"blockName":"core\/paragraph","attrs":{"textColor":"medium-grey","align":"","content":"Text adapted from an article published in Le Cahier des Ponts n\u00b01, L'intelligence artificielle, March 2019.","dropCap":false,"placeholder":"","direction":"","lock":[],"metadata":[],"className":"has-medium-grey-color has-text-color","style":"","backgroundColor":"","gradient":"","fontSize":"","fontFamily":"","borderColor":"","anchor":""},"innerBlocks":[],"innerHTML":"\n<p class=\"has-medium-grey-color has-text-color\"><strong>Text adapted from an article published in <a rel=\"noreferrer noopener\" href=\"https:\/\/ecoledesponts.fr\/sites\/ecoledesponts.fr\/files\/documents\/cdp_num1_ia.pdf\" target=\"_blank\">Le Cahier des Ponts n\u00b01<\/a>, L'intelligence artificielle, March 2019<\/strong>.<\/p>\n","innerContent":["\n<p class=\"has-medium-grey-color has-text-color\"><strong>Text adapted from an article published in <a rel=\"noreferrer noopener\" href=\"https:\/\/ecoledesponts.fr\/sites\/ecoledesponts.fr\/files\/documents\/cdp_num1_ia.pdf\" target=\"_blank\">Le Cahier des Ponts n\u00b01<\/a>, L'intelligence artificielle, March 2019<\/strong>.<\/p>\n"],"rendered":"\n<p class=\"has-medium-grey-color has-text-color\"><strong>Text adapted from an article published in <a rel=\"noreferrer noopener\" href=\"https:\/\/ecoledesponts.fr\/sites\/ecoledesponts.fr\/files\/documents\/cdp_num1_ia.pdf\" target=\"_blank\">Le Cahier des Ponts n\u00b01<\/a>, L'intelligence artificielle, March 2019<\/strong>.<\/p>\n"}],"seo":{"title":"Artificial Intelligence: a scientific and technological revolution"},"media":{"img":"<img width=\"150\" height=\"150\" src=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2022\/06\/Cdp_Numero_1_02_Carte_blanche_Fig_01_Grande_illu-150x150-2.jpg\" class=\"attachment-full size-full\" alt=\"\" decoding=\"async\" loading=\"lazy\" srcset=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2022\/06\/Cdp_Numero_1_02_Carte_blanche_Fig_01_Grande_illu-150x150-2.jpg 150w, https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2022\/06\/Cdp_Numero_1_02_Carte_blanche_Fig_01_Grande_illu-150x150-2-60x60.jpg 60w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/>","src":"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2022\/06\/Cdp_Numero_1_02_Carte_blanche_Fig_01_Grande_illu-150x150-2.jpg"},"url":"\/en\/articles\/artificial-intelligence-a-scientific-and-technological-revolution\/","related":{"post":[],"author":[{"title":"Axel Parmentier","url":"\/en\/authors\/axel-parmentier\/","id":"946","media":"<img width=\"60\" height=\"60\" src=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2022\/11\/Axel_Parmentier-60x60.png\" class=\"attachment-author-thumb size-author-thumb wp-post-image\" alt=\"\" decoding=\"async\" loading=\"lazy\" srcset=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2022\/11\/Axel_Parmentier-60x60.png 60w, https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2022\/11\/Axel_Parmentier-150x150.png 150w\" sizes=\"auto, (max-width: 60px) 100vw, 60px\" \/>","slug":"axel-parmentier"}],"subject":[{"title":"Digital Technology, Modeling &#038; Artificial Intelligence","url":"\/en\/subjects\/digital-technology-modeling-artificial-intelligence\/","id":"690","media":"<img width=\"1920\" height=\"1080\" src=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2022\/11\/Ecole-des-ponts-webmagazine-numerique.jpg\" class=\"attachment- size- wp-post-image\" alt=\"\" decoding=\"async\" loading=\"lazy\" srcset=\"https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2022\/11\/Ecole-des-ponts-webmagazine-numerique.jpg 1920w, https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2022\/11\/Ecole-des-ponts-webmagazine-numerique-300x169.jpg 300w, https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2022\/11\/Ecole-des-ponts-webmagazine-numerique-1024x576.jpg 1024w, https:\/\/ingenius.ecoledesponts.fr\/wp-content\/uploads\/2022\/11\/Ecole-des-ponts-webmagazine-numerique-768x432.jpg 768w\" sizes=\"auto, (max-width: 1920px) 100vw, 1920px\" \/>","slug":"digital-technology-modeling-artificial-intelligence"}],"category":[{"title":"Articles","url":"\/en\/articles\/category\/articles\/","id":"1716","media":"","slug":"articles","_related_post_type":""}],"folder":[]},"translated":"https:\/\/ingenius.ecoledesponts.fr\/articles\/lintelligence-artificielle-une-revolution-scientifique-et-technologique\/","icon":"icon-article","duration":"6","custom_excerpt":"Artificial intelligence (AI) has been around for decades, but recently it has experienced significant advances in a growing number of industries. From the placement of advertising spots on a television channel to airline company schedules and even the optimal management of hydroelectric dam valleys, AI research has a bright future ahead of it.","duration_type":"","_links":{"self":[{"href":"https:\/\/ingenius.ecoledesponts.fr\/en\/wp-json\/wp\/v2\/posts\/1769","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/ingenius.ecoledesponts.fr\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/ingenius.ecoledesponts.fr\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/ingenius.ecoledesponts.fr\/en\/wp-json\/wp\/v2\/users\/6"}],"replies":[{"embeddable":true,"href":"https:\/\/ingenius.ecoledesponts.fr\/en\/wp-json\/wp\/v2\/comments?post=1769"}],"version-history":[{"count":5,"href":"https:\/\/ingenius.ecoledesponts.fr\/en\/wp-json\/wp\/v2\/posts\/1769\/revisions"}],"predecessor-version":[{"id":8931,"href":"https:\/\/ingenius.ecoledesponts.fr\/en\/wp-json\/wp\/v2\/posts\/1769\/revisions\/8931"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/ingenius.ecoledesponts.fr\/en\/wp-json\/wp\/v2\/media\/1464"}],"wp:attachment":[{"href":"https:\/\/ingenius.ecoledesponts.fr\/en\/wp-json\/wp\/v2\/media?parent=1769"}],"wp:term":[{"taxonomy":"article-types","embeddable":true,"href":"https:\/\/ingenius.ecoledesponts.fr\/en\/wp-json\/wp\/v2\/article-types?post=1769"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}