Worth: $92.00
(as of Nov 23, 2025 01:29:40 UTC – Particulars)
The primary complete introduction to Multi-Agent Reinforcement Studying (MARL), overlaying MARL’s fashions, resolution ideas, algorithmic concepts, technical challenges, and fashionable approaches.
Multi-Agent Reinforcement Studying (MARL), an space of machine studying during which a collective of brokers study to optimally work together in a shared setting, boasts a rising array of purposes in fashionable life, from autonomous driving and multi-robot factories to automated buying and selling and power community administration. This textual content supplies a lucid and rigorous introduction to the fashions, resolution ideas, algorithmic concepts, technical challenges, and fashionable approaches in MARL. The e-book first introduces the sector’s foundations, together with fundamentals of reinforcement studying idea and algorithms, interactive sport fashions, completely different resolution ideas for video games, and the algorithmic concepts underpinning MARL analysis. It then particulars modern MARL algorithms which leverage deep studying methods, overlaying concepts corresponding to centralized coaching with decentralized execution, worth decomposition, parameter sharing, and self-play. The e-book comes with its personal MARL codebase written in Python, containing implementations of MARL algorithms which are self-contained and straightforward to learn. Technical content material is defined in easy-to-understand language and illustrated with intensive examples, illuminating MARL for newcomers whereas providing high-level insights for extra superior readers.
First textbook to introduce the foundations and purposes of MARL, written by consultants within the fieldIntegrates reinforcement studying, deep studying, and sport theoryPractical focus covers concerns for operating experiments and describes environments for testing MARL algorithmsExplains complicated ideas in clear and easy languageClassroom-tested, accessible strategy appropriate for graduate college students and professionals throughout pc science, synthetic intelligence, and robotics Assets embody code and slides
Writer : The MIT Press
Publication date : Dec 17 2024
Language : English
Print size : 396 pages
ISBN-10 : 0262049376
ISBN-13 : 978-0262049375
Merchandise weight : 760 g
Dimensions : 16.03 x 2.87 x 23.65 cm
Greatest Sellers Rank: #581,472 in Books (See Prime 100 in Books) #192 in A.I. Neural Networks #241 in Pc Science Textbooks #518 in Pc Algorithms
Buyer Critiques: 4.6 4.6 out of 5 stars 9 rankings var dpAcrHasRegisteredArcLinkClickAction; P.when(‘A’, ‘prepared’).execute(operate(A) { if (dpAcrHasRegisteredArcLinkClickAction !== true) { dpAcrHasRegisteredArcLinkClickAction = true; A.declarative( ‘acrLink-click-metrics’, ‘click on’, { “allowLinkDefault”: true }, operate (occasion) { if (window.ue) 0) + 1); } ); } }); P.when(‘A’, ‘cf’).execute(operate(A) { A.declarative(‘acrStarsLink-click-metrics’, ‘click on’, { “allowLinkDefault” : true }, operate(occasion){ if(window.ue) }); });
