Evaluating Reinforcement Learning algorithms forLunarLander-v2: A Comparative Analysis

Evaluating Reinforcement Learning algorithms forLunarLander-v2: A Comparative Analysis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Evaluating Reinforcement Learning algorithms forLunarLander-v2: A Comparative Analysis Anurag Awasthi This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5939959/v2 This work is licensed under a CC BY 4.0 License Status: Posted Version 2 posted You are reading this latest preprint version Show more versions Abstract This study investigates the performance and optimization strategies of advanced reinforcement learning algorithms: Deep Q-Network (DQN), Double Deep Q-Network (DDQN), Deep Deterministic Policy Gradient (DDPG) and Proximal Policy Optimization (PPO) to solve the LunarLander-v2 control task. Each algorithm is analyzed on the basis of its design principles, training stability, and adaptability to complex action and state spaces. DDQN addresses overestimation biases in DQN by leveraging dual networks for more accurate value estimation. DDPG incorporates an actor-critic architecture to excel in continuous control tasks, while PPO introduces a policy gradient approach with clipped updates to balance exploration and exploitation. Hyperparameter tuning, including learning rates, discount factors, and batch sizes, is explored to optimize the performance of each method. The results demonstrate significant improvements in stability, learning efficiency, and final scores as training progresses. Comparative insights highlight the trade-offs between sample efficiency, stability, and computational overhead across these algorithms, providing valuable guidance for reinforcement learning applications in dynamic control environments. Artificial Intelligence and Machine Learning Machine Learning Reinforcement Learning Full Text Additional Declarations The authors declare no competing interests. Cite Share Download PDF Status: Posted Version 2 posted You are reading this latest preprint version Show more versions Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5939959","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":410792173,"identity":"3c996a79-fa7a-4b08-b08e-7b1d0bf7ac47","order_by":0,"name":"Anurag Awasthi","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA30lEQVRIiWNgGAWjYDADfjBZQIoWyQYQaUCKFoMDxGrhn91j9uHHL7vEzecXH3z4xcCGgb+9OwGvFok7Z4xn9vYlJ2678SzZWMYgjUHizNkN+K25kWPMwNvDnLvtxhkzaQmDwwwGErn4tcgDtTD+7anP3TwDrOU/YS0GQC3MPD8O527g7zGT/AAMBIJaDG+kFTPLNhyvn3GDLdmYwSCZh6Bf5G4kb2Z886famL//8MGHPyrs5Pjbewl4HwQY24CERAIDMw8DAw9h5WDwB4j5DzAw/iBS/SgYBaNgFIwsAACYhUn1kUG1sgAAAABJRU5ErkJggg==","orcid":"","institution":"Google","correspondingAuthor":true,"prefix":"","firstName":"Anurag","middleName":"","lastName":"Awasthi","suffix":""}],"badges":[],"createdAt":"2025-02-01 05:24:48","currentVersionCode":2,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-5939959/v2","doiUrl":"https://doi.org/10.21203/rs.3.rs-5939959/v2","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":80653407,"identity":"7ab1aabc-82d0-43f2-8366-490ae44e18ba","added_by":"auto","created_at":"2025-04-15 15:13:39","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":375199,"visible":true,"origin":"","legend":"","description":"","filename":"EvaluatingReinforcementLearningAlgorithmsforLunarLanderv2AComparativeAnalysisofDQNDDQNDDPGandPPO14.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5939959/v2_covered_b0beacb0-534e-4c72-8a6c-7954fdc6a3ed.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eEvaluating Reinforcement Learning algorithms forLunarLander-v2: A Comparative Analysis\u003c/p\u003e","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"Google (United States)","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Machine Learning, Reinforcement Learning","lastPublishedDoi":"10.21203/rs.3.rs-5939959/v2","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5939959/v2","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis study investigates the performance and optimization strategies of advanced reinforcement learning algorithms: Deep Q-Network (DQN), Double Deep Q-Network (DDQN), Deep Deterministic Policy Gradient (DDPG) and Proximal Policy Optimization (PPO) to solve the LunarLander-v2 control task. Each algorithm is analyzed on the basis of its design principles, training stability, and adaptability to complex action and state spaces. DDQN addresses overestimation biases in DQN by leveraging dual networks for more accurate value estimation. DDPG incorporates an actor-critic architecture to excel in continuous control tasks, while PPO introduces a policy gradient approach with clipped updates to balance exploration and exploitation. Hyperparameter tuning, including learning rates, discount factors, and batch sizes, is explored to optimize the performance of each method. The results demonstrate significant improvements in stability, learning efficiency, and final scores as training progresses. Comparative insights highlight the trade-offs between sample efficiency, stability, and computational overhead across these algorithms, providing valuable guidance for reinforcement learning applications in dynamic control environments.\u003c/p\u003e","manuscriptTitle":"Evaluating Reinforcement Learning algorithms forLunarLander-v2: A Comparative Analysis","msid":"","msnumber":"","nonDraftVersions":[{"code":2,"date":"2025-04-15 15:05:34","doi":"10.21203/rs.3.rs-5939959/v2","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}},{"code":1,"date":"2025-02-04 02:23:54","doi":"10.21203/rs.3.rs-5939959/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"b3fb3bae-ddf9-4c69-9294-083f2c88d107","owner":[],"postedDate":"April 15th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":47203941,"name":"Artificial Intelligence and Machine Learning"}],"tags":[],"updatedAt":"2025-02-04T02:23:54+00:00","versionOfRecord":[],"versionCreatedAt":"2025-04-15 15:05:34","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v2","identity":"rs-5939959","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5939959","identity":"rs-5939959","version":["v2"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}