Chapter I  Artificial Intelligence Introduction What Is AI?The Foundations of Artificial IntelligenceThe History of Artificial IntelligenceThe State of the ArtRisks and Benefits of AISummaryBibliographical and Historical NotesIntelligent Agents Agents and EnvironmentsGood Behavior: The Concept of RationalityThe Nature of EnvironmentsThe Structure of AgentsSummaryBibliographical and Historical NotesChapter II  Problem Solving Solving Problems by Searching Problem-Solving AgentsExample ProblemsSearch AlgorithmsUninformed Search StrategiesInformed (Heuristic) Search StrategiesHeuristic FunctionsSummaryBibliographical and Historical NotesSearch in Complex Environments Local Search and Optimization ProblemsLocal Search in Continuous SpacesSearch with Nondeterministic ActionsSearch in Partially Observable EnvironmentsOnline Search Agents and Unknown EnvironmentsSummaryBibliographical and Historical NotesConstraint Satisfaction Problems Defining Constraint Satisfaction ProblemsConstraint Propagation: Inference in CSPsBacktracking Search for CSPsLocal Search for CSPsThe Structure of ProblemsSummaryBibliographical and Historical NotesAdversarial Search and Games Game TheoryOptimal Decisions in GamesHeuristic Alpha--Beta Tree SearchMonte Carlo Tree SearchStochastic GamesPartially Observable GamesLimitations of Game Search AlgorithmsSummaryBibliographical and Historical NotesChapter III  Knowledge, Reasoning and Planning Logical Agents Knowledge-Based AgentsThe Wumpus WorldLogicPropositional Logic: A Very Simple LogicPropositional Theorem ProvingEffective Propositional Model CheckingAgents Based on Propositional LogicSummaryBibliographical and Historical NotesFirst-Order Logic Representation RevisitedSyntax and Semantics of First-Order LogicUsing First-Order LogicKnowledge Engineering in First-Order LogicSummaryBibliographical and Historical NotesInference in First-Order Logic Propositional vs. First-Order InferenceUnification and First-Order InferenceForward ChainingBackward ChainingResolutionSummaryBibliographical and Historical NotesKnowledge Representation Ontological EngineeringCategories and ObjectsEventsMental Objects and Modal Logicfor CategoriesReasoning with Default InformationSummaryBibliographical and Historical NotesAutomated Planning Definition of Classical PlanningAlgorithms for Classical PlanningHeuristics for PlanningHierarchical PlanningPlanning and Acting in Nondeterministic DomainsTime, Schedules, and ResourcesAnalysis of Planning ApproachesSummaryBibliographical and Historical NotesChapter IV  Uncertain Knowledge and Reasoning Quantifying Uncertainty Acting under UncertaintyBasic Probability NotationInference Using Full Joint DistributionsIndependence 12.5 Bayes' Rule and Its UseNaive Bayes ModelsThe Wumpus World RevisitedSummaryBibliographical and Historical NotesProbabilistic Reasoning Representing Knowledge in an Uncertain DomainThe Semantics of Bayesian NetworksExact Inference in Bayesian NetworksApproximate Inference for Bayesian NetworksCausal NetworksSummaryBibliographical and Historical NotesProbabilistic Reasoning over Time Time and UncertaintyInference in Temporal ModelsHidden Markov ModelsKalman FiltersDynamic Bayesian NetworksSummaryBibliographical and Historical NotesMaking Simple Decisions Combining Beliefs and Desires under UncertaintyThe Basis of Utility TheoryUtility FunctionsMultiattribute Utility FunctionsDecision NetworksThe Value of InformationUnknown PreferencesSummaryBibliographical and Historical NotesMaking Complex Decisions Sequential Decision ProblemsAlgorithms for MDPsBandit ProblemsPartially Observable MDPsAlgorithms for Solving POMDPsSummaryBibliographical and Historical NotesMultiagent Decision Making Properties of Multiagent EnvironmentsNon-Cooperative Game TheoryCooperative Game TheoryMaking Collective DecisionsSummaryBibliographical and Historical NotesProbabilistic Programming Relational Probability ModelsOpen-Universe Probability ModelsKeeping Track of a Complex WorldPrograms as Probability ModelsSummaryBibliographical and Historical NotesChapter V  Machine Learning Learning from Examples Forms of LeamingSupervised Learning .Learning Decision Trees .Model Selection and OptimizationThe Theory of LearningLinear Regression and ClassificationNonparametric ModelsEnsemble LearningDeveloping Machine Learning SystenSummaryBibliographical and Historical NotesKnowledge in Learning A Logical Formulation of LearningKnowledge in LearningExmplanation-Based LeaeningLearning Using Relevance InformationInductive Logic ProgrammingSummaryBibliographical and Historical NotesLearning Probabilistic Models Statistical LearningLearning with Complete DataLearning with Hidden Variables: The EM AlgorithmSummaryBibliographical and Historical NotesDeep Learning Simple Feedforward NetworksComputation Graphs for Deep LearningConvolutional NetworksLearning AlgorithmsGeneralizationRecurrent Neural NetworksUnsupervised Learning and Transfer LearningApplicationsSummaryBibliographical and Historical NotesReinforcement Learning Learning from RewardsPassive Reinforcement LearningActive Reinforcement LearningGeneralization in Reinforcement LearningPolicy SearchApprenticeship and Inverse Reinforcement LeamingApplications of Reinforcement LearningSummaryBibliographical and Historical NotesChapter VI  Communicating, perceiving, and acting Natural Language Processing Language ModelsGrammarParsingAugmented GrammarsComplications of Real Natural LanguagrNatural Language TasksSummaryBibliographical and Historical NotesDeep Learning for Natural Language Processing Word EmbeddingsRecurrent Neural Networks for NLPSequence-to-Sequence ModelsThe Transformer ArchitecturePretraining and Transfer LearningState of the artSummaryBibliographical and Historical NotesRobotics RobotsRobot HardwareWhat kind of problem is robotics solving?Robotic PerceptionPlanning and ControlPlanning Uncertain MovementsReinforcement Laming in RoboticsHumans and RobotsAlternative Robotic FrameworksApplication DomainsSummaryBibliographical and Historical NotesComputer Vision IntroductionImage FormationSimple Image FeaturesClassifying ImagesDetecting ObjectsThe 3D WorldUsing Computer VisionSummaryBibliographical and Historical NotesChapter VII  Conclusions Philosophy, Ethics, and Safety of Al The Limits of AlCan Machines Really Think?The Ethics of AlSummaryBibliographical and Historical NotesThe Future of AI Al ComponentsAl Architectures A Mathematical Background A.1 Complexity Analysis and O0 NotationA.2 Vectors, Matrices, and Linear AlgebraA.3 Probability DistributionsBibliographical and Historical Notes   B Notes on Languages and Algorithms B.1 Defining Languages with Backus-Naur Form (BNF)B.2 Describing Algorithms with PseudocodeB.3 Online Supplemental Material   Bibliography Index

Hallmark features of this title Offer the most comprehensive and accessible introduction to the theory and practice of AI In-depth coverage of both basic and advanced topics provides students with a basic understanding of the discipline without compromising complexity and depth.Non-technical learning material makes content more accessible, introducing major concepts before going into mathematical algorithmic details. Teach up-to-date material in a more unified manner according to latest technologies Unified approach to AI showcases how the various sub-fields of AI fit together.Flexible format offers adaptable text for varying instructors' preferences. A comprehensive index and extensive bibliography support student learning, covering a wide range of topics relevant to the content.

New and updated features of this titleUpdated and new content reflects recent technological advancements and applications in the field. New chapters feature expanded coverage of probabilistic programming (Ch. 18); multi-agent decision making (Ch. 17 with Michael Wooldridge); deep learning (Ch.22 with Ian Goodfellow); and deep learning for natural language processing (Ch. 25 with Jacob Devlin and Mei-Wing Chang).Significantly updated material on robotics includes robots that interact with humans and the application of reinforcement learning to robotics.Revised coverage of computer vision, natural language understanding, and speech recognition reflects the impact of deep learning methods on these fields.The Author-Maintained Website at http://aima.cs.berkeley.edu/ includes text-related comments and discussions, exercises, an online code repository, Instructor Resources, and more.New sections include content on current technologies, as well as the ethical aspects and values of the discipline. New sections include content on topics, such as causality by Judea Pearl, and the Monte Carlo search for games and robotics.New sections discuss transfer learning for deep learning in general and for natural language.New sections include content on privacy, fairness, future of work, and safe AI.