{"id":5854,"date":"2023-03-03T14:33:16","date_gmt":"2023-03-03T14:33:16","guid":{"rendered":"https:\/\/aaep.org.ar\/?p=5854"},"modified":"2023-03-03T14:33:19","modified_gmt":"2023-03-03T14:33:19","slug":"q-learning-algorithms-in-a-hotelling-model","status":"publish","type":"post","link":"https:\/\/aaep.org.ar\/?p=5854","title":{"rendered":"Q-Learning algorithms in a Hotelling model"},"content":{"rendered":"<iframe loading=\"lazy\" class=\"wonderplugin-pdf-iframe\" src=\"https:\/\/aaep.org.ar\/wp-content\/plugins\/wonderplugin-pdf-embed\/pdfjslight\/web\/viewer.html?v=2&file=https:\/\/aaep.org.ar\/works\/works2022\/4587.pdf\" width=\"100%\" height=\"600px\" style=\"border:0;\"><\/iframe>\n","protected":false},"excerpt":{"rendered":"<p>What if Q-Learning algorithms set not only prices but also the degree of differentiation between them? In this paper, I tackle this question by analyzing the competition between two Q-Learning algorithms in a Hotelling setting. I find that most of the simulations converge to a Nash Equilibrium where the algorithms are playing non-competitive strategies. In most simulations, they optimally learn not to differentiate each other and to set a collusive price. An underlying deviation and punishment scheme sustains this implicit agreement. The results are robust to the enlargement of the action space and the introduction of relocalization costs.<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"_coblocks_attr":"","_coblocks_dimensions":"","_coblocks_responsive_height":"","_coblocks_accordion_ie_support":"","footnotes":""},"categories":[29],"tags":[30],"class_list":["post-5854","post","type-post","status-publish","format-standard","hentry","category-anales","tag-aaep-anales-2022"],"acf":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v27.5 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>Q-Learning algorithms in a Hotelling model - Asociaci\u00f3n Argentina de Econom\u00eda Pol\u00edtica<\/title>\n<meta name=\"robots\" content=\"noindex, follow\" \/>\n<meta property=\"og:locale\" content=\"es_ES\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Q-Learning algorithms in a Hotelling model - Asociaci\u00f3n Argentina de Econom\u00eda Pol\u00edtica\" \/>\n<meta property=\"og:description\" content=\"What if Q-Learning algorithms set not only prices but also the degree of differentiation between them? In this paper, I tackle this question by analyzing the competition between two Q-Learning algorithms in a Hotelling setting. I find that most of the simulations converge to a Nash Equilibrium where the algorithms are playing non-competitive strategies. In most simulations, they optimally learn not to differentiate each other and to set a collusive price. An underlying deviation and punishment scheme sustains this implicit agreement. The results are robust to the enlargement of the action space and the introduction of relocalization costs.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/aaep.org.ar\/?p=5854\" \/>\n<meta property=\"og:site_name\" content=\"Asociaci\u00f3n Argentina de Econom\u00eda Pol\u00edtica\" \/>\n<meta property=\"article:published_time\" content=\"2023-03-03T14:33:16+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2023-03-03T14:33:19+00:00\" \/>\n<meta name=\"author\" content=\"aaepolitica\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@aaepoficial\" \/>\n<meta name=\"twitter:site\" content=\"@aaepoficial\" \/>\n<meta name=\"twitter:label1\" content=\"Escrito por\" \/>\n\t<meta name=\"twitter:data1\" content=\"aaepolitica\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\\\/\\\/aaep.org.ar\\\/?p=5854#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/aaep.org.ar\\\/?p=5854\"},\"author\":{\"name\":\"aaepolitica\",\"@id\":\"https:\\\/\\\/aaep.org.ar\\\/#\\\/schema\\\/person\\\/7150b747ba20875259c74b2cd177316c\"},\"headline\":\"Q-Learning algorithms in a Hotelling model\",\"datePublished\":\"2023-03-03T14:33:16+00:00\",\"dateModified\":\"2023-03-03T14:33:19+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/aaep.org.ar\\\/?p=5854\"},\"wordCount\":21,\"publisher\":{\"@id\":\"https:\\\/\\\/aaep.org.ar\\\/#organization\"},\"keywords\":[\"AAEP Anales 2022\"],\"articleSection\":[\"Anales\"],\"inLanguage\":\"es\"},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/aaep.org.ar\\\/?p=5854\",\"url\":\"https:\\\/\\\/aaep.org.ar\\\/?p=5854\",\"name\":\"Q-Learning algorithms in a Hotelling model - Asociaci\u00f3n Argentina de Econom\u00eda Pol\u00edtica\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/aaep.org.ar\\\/#website\"},\"datePublished\":\"2023-03-03T14:33:16+00:00\",\"dateModified\":\"2023-03-03T14:33:19+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/aaep.org.ar\\\/?p=5854#breadcrumb\"},\"inLanguage\":\"es\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/aaep.org.ar\\\/?p=5854\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/aaep.org.ar\\\/?p=5854#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Portada\",\"item\":\"https:\\\/\\\/aaep.org.ar\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Q-Learning algorithms in a Hotelling model\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/aaep.org.ar\\\/#website\",\"url\":\"https:\\\/\\\/aaep.org.ar\\\/\",\"name\":\"Asociaci\u00f3n Argentina de Econom\u00eda Pol\u00edtica\",\"description\":\"Asociaci\u00f3n Argentina de Econom\u00eda Pol\u00edtica\",\"publisher\":{\"@id\":\"https:\\\/\\\/aaep.org.ar\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/aaep.org.ar\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"es\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/aaep.org.ar\\\/#organization\",\"name\":\"Asociaci\u00f3n Argentina de Econom\u00eda Pol\u00edtica\",\"url\":\"https:\\\/\\\/aaep.org.ar\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"es\",\"@id\":\"https:\\\/\\\/aaep.org.ar\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/aaep.org.ar\\\/wp-content\\\/uploads\\\/2022\\\/05\\\/cropped-aaep_logo.png\",\"contentUrl\":\"https:\\\/\\\/aaep.org.ar\\\/wp-content\\\/uploads\\\/2022\\\/05\\\/cropped-aaep_logo.png\",\"width\":74,\"height\":49,\"caption\":\"Asociaci\u00f3n Argentina de Econom\u00eda Pol\u00edtica\"},\"image\":{\"@id\":\"https:\\\/\\\/aaep.org.ar\\\/#\\\/schema\\\/logo\\\/image\\\/\"},\"sameAs\":[\"https:\\\/\\\/x.com\\\/aaepoficial\",\"https:\\\/\\\/www.instagram.com\\\/aaepoficial\\\/\"]},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/aaep.org.ar\\\/#\\\/schema\\\/person\\\/7150b747ba20875259c74b2cd177316c\",\"name\":\"aaepolitica\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"es\",\"@id\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/b4c835a24b26f7310194c9c19eee26567492d7a70e27fc0adad2d6b1983478c9?s=96&d=mm&r=g\",\"url\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/b4c835a24b26f7310194c9c19eee26567492d7a70e27fc0adad2d6b1983478c9?s=96&d=mm&r=g\",\"contentUrl\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/b4c835a24b26f7310194c9c19eee26567492d7a70e27fc0adad2d6b1983478c9?s=96&d=mm&r=g\",\"caption\":\"aaepolitica\"},\"sameAs\":[\"https:\\\/\\\/aaep.org.ar\"]}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Q-Learning algorithms in a Hotelling model - Asociaci\u00f3n Argentina de Econom\u00eda Pol\u00edtica","robots":{"index":"noindex","follow":"follow"},"og_locale":"es_ES","og_type":"article","og_title":"Q-Learning algorithms in a Hotelling model - Asociaci\u00f3n Argentina de Econom\u00eda Pol\u00edtica","og_description":"What if Q-Learning algorithms set not only prices but also the degree of differentiation between them? In this paper, I tackle this question by analyzing the competition between two Q-Learning algorithms in a Hotelling setting. I find that most of the simulations converge to a Nash Equilibrium where the algorithms are playing non-competitive strategies. In most simulations, they optimally learn not to differentiate each other and to set a collusive price. An underlying deviation and punishment scheme sustains this implicit agreement. The results are robust to the enlargement of the action space and the introduction of relocalization costs.","og_url":"https:\/\/aaep.org.ar\/?p=5854","og_site_name":"Asociaci\u00f3n Argentina de Econom\u00eda Pol\u00edtica","article_published_time":"2023-03-03T14:33:16+00:00","article_modified_time":"2023-03-03T14:33:19+00:00","author":"aaepolitica","twitter_card":"summary_large_image","twitter_creator":"@aaepoficial","twitter_site":"@aaepoficial","twitter_misc":{"Escrito por":"aaepolitica"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/aaep.org.ar\/?p=5854#article","isPartOf":{"@id":"https:\/\/aaep.org.ar\/?p=5854"},"author":{"name":"aaepolitica","@id":"https:\/\/aaep.org.ar\/#\/schema\/person\/7150b747ba20875259c74b2cd177316c"},"headline":"Q-Learning algorithms in a Hotelling model","datePublished":"2023-03-03T14:33:16+00:00","dateModified":"2023-03-03T14:33:19+00:00","mainEntityOfPage":{"@id":"https:\/\/aaep.org.ar\/?p=5854"},"wordCount":21,"publisher":{"@id":"https:\/\/aaep.org.ar\/#organization"},"keywords":["AAEP Anales 2022"],"articleSection":["Anales"],"inLanguage":"es"},{"@type":"WebPage","@id":"https:\/\/aaep.org.ar\/?p=5854","url":"https:\/\/aaep.org.ar\/?p=5854","name":"Q-Learning algorithms in a Hotelling model - Asociaci\u00f3n Argentina de Econom\u00eda Pol\u00edtica","isPartOf":{"@id":"https:\/\/aaep.org.ar\/#website"},"datePublished":"2023-03-03T14:33:16+00:00","dateModified":"2023-03-03T14:33:19+00:00","breadcrumb":{"@id":"https:\/\/aaep.org.ar\/?p=5854#breadcrumb"},"inLanguage":"es","potentialAction":[{"@type":"ReadAction","target":["https:\/\/aaep.org.ar\/?p=5854"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/aaep.org.ar\/?p=5854#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Portada","item":"https:\/\/aaep.org.ar\/"},{"@type":"ListItem","position":2,"name":"Q-Learning algorithms in a Hotelling model"}]},{"@type":"WebSite","@id":"https:\/\/aaep.org.ar\/#website","url":"https:\/\/aaep.org.ar\/","name":"Asociaci\u00f3n Argentina de Econom\u00eda Pol\u00edtica","description":"Asociaci\u00f3n Argentina de Econom\u00eda Pol\u00edtica","publisher":{"@id":"https:\/\/aaep.org.ar\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/aaep.org.ar\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"es"},{"@type":"Organization","@id":"https:\/\/aaep.org.ar\/#organization","name":"Asociaci\u00f3n Argentina de Econom\u00eda Pol\u00edtica","url":"https:\/\/aaep.org.ar\/","logo":{"@type":"ImageObject","inLanguage":"es","@id":"https:\/\/aaep.org.ar\/#\/schema\/logo\/image\/","url":"https:\/\/aaep.org.ar\/wp-content\/uploads\/2022\/05\/cropped-aaep_logo.png","contentUrl":"https:\/\/aaep.org.ar\/wp-content\/uploads\/2022\/05\/cropped-aaep_logo.png","width":74,"height":49,"caption":"Asociaci\u00f3n Argentina de Econom\u00eda Pol\u00edtica"},"image":{"@id":"https:\/\/aaep.org.ar\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/x.com\/aaepoficial","https:\/\/www.instagram.com\/aaepoficial\/"]},{"@type":"Person","@id":"https:\/\/aaep.org.ar\/#\/schema\/person\/7150b747ba20875259c74b2cd177316c","name":"aaepolitica","image":{"@type":"ImageObject","inLanguage":"es","@id":"https:\/\/secure.gravatar.com\/avatar\/b4c835a24b26f7310194c9c19eee26567492d7a70e27fc0adad2d6b1983478c9?s=96&d=mm&r=g","url":"https:\/\/secure.gravatar.com\/avatar\/b4c835a24b26f7310194c9c19eee26567492d7a70e27fc0adad2d6b1983478c9?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/b4c835a24b26f7310194c9c19eee26567492d7a70e27fc0adad2d6b1983478c9?s=96&d=mm&r=g","caption":"aaepolitica"},"sameAs":["https:\/\/aaep.org.ar"]}]}},"_links":{"self":[{"href":"https:\/\/aaep.org.ar\/index.php?rest_route=\/wp\/v2\/posts\/5854","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/aaep.org.ar\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/aaep.org.ar\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/aaep.org.ar\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/aaep.org.ar\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=5854"}],"version-history":[{"count":2,"href":"https:\/\/aaep.org.ar\/index.php?rest_route=\/wp\/v2\/posts\/5854\/revisions"}],"predecessor-version":[{"id":5856,"href":"https:\/\/aaep.org.ar\/index.php?rest_route=\/wp\/v2\/posts\/5854\/revisions\/5856"}],"wp:attachment":[{"href":"https:\/\/aaep.org.ar\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=5854"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/aaep.org.ar\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=5854"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/aaep.org.ar\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=5854"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}