This preview shows page 1-2-3-4-5-6-7-8-9-63-64-65-66-67-68-69-70-71-127-128-129-130-131-132-133-134-135 out of 135 pages.
Overview and summaryActing Humanly: The Turing TestWhat would a computer need to pass the Turing test?Slide 4What is an (Intelligent) Agent?Environment typesSlide 7Agent typesReflex agentsReflex agents w/ stateGoal-based agentsUtility-based agentsHow can we design & implement agents?Problem-Solving AgentProblem typesSearch algorithmsImplementation of search algorithmsEncapsulating state information in nodesComplexityWhy is exponential complexity “hard”?Landau symbolsPolynomial-time hierarchySearch strategiesSlide 24Example: Traveling from Arad To BucharestBreadth-first searchSlide 27Slide 28Uniform-cost searchSlide 30Slide 31Depth-first searchSlide 33Slide 34PowerPoint PresentationSlide 36Slide 37Slide 38Slide 39Slide 40Slide 41Slide 42Informed search: Best-first searchGreedy searchA* searchComparing uninformed search strategiesSlide 47Iterative improvementHill climbing (or gradient ascent/descent)Simulated AnnealingSimulated annealing algorithmNote on simulated annealing: limit casesIs search applicable to game playing?Searching for the next moveThe minimax algorithmminimax = maximum of the minimum- pruning: search cutoff- pruning: exampleSlide 59Slide 60Slide 61Nondeterministic games: the element of chanceSlide 63Summary on gamesKnowledge-Based AgentGeneric knowledge-based agentLogic in generalTypes of logicEntailmentInferenceValidity and satisfiabilityPropositional logic: semanticsPropositional inference: normal formsProof methodsInference rulesLimitations of Propositional LogicFirst-order logic (FOL)Universal quantification (for all): Existential quantification (there exists): Properties of quantifiersExample sentencesHigher-order logic?Using the FOL Knowledge BaseWumpus world, FOL Knowledge BaseDeducing hidden propertiesSituation calculusDescribing actionsDescribing actions (cont’d)PlanningGenerating action sequencesSummary on FOLKnowledge EngineerKnowledge engineering vs. programmingTowards a general ontologyInference in First-Order LogicProofsGeneralized Modus Ponens (GMP)Forward chainingBackward chainingResolutionResolution inference ruleResolution proofLogical reasoning systemsSlide 104Membership functions: S-functionMembership functions: P-FunctionLinguistic HedgesFuzzy set operatorsFuzzy inference overviewCLIPS Inference cycleWhat we have so farSearch vs. planningTypes of plannersA Simple Planning AgentSTRIPS operatorsPartially ordered plansPlanPOP algorithm sketchPOP algorithm (cont.)Some problems remain…Computer PerceptionPerception for what?Image analysis/Computer visionVisual AttentionPedestrian recognitionSlide 126More robot examplesWarren McCulloch and Walter Pitts (1943)Leaky Integrator NeuronLeaky Integrator ModelHopfield Networks“Energy” of a Neural NetworkSelf-Organizing Feature MapsExample: face recognitionAssociative MemoriesCS 561, Session 281Overview and summaryWe have discussed…- What AI and intelligent agents are- How to develop AI systems - How to solve problems using search- How to play games as an application/extension of search- How to build basic agents that reason logically,using propositional logic- How to write more powerful logic statements with first-order logic- How to properly engineer a knowledge base- How to reason logically using first-order logic inference- Examples of logical reasoning systems, such as theorem provers- How to plan- Expert systems- What challenges remainCS 561, Session 282Acting Humanly: The Turing Test•Alan Turing's 1950 article Computing Machinery and Intelligence discussed conditions for considering a machine to be intelligent•“Can machines think?” “Can machines behave intelligently?”•The Turing test (The Imitation Game): Operational definition of intelligence.•Computer needs to posses: Natural language processing, Knowledge representation, Automated reasoning, and Machine learningCS 561, Session 283What would a computer need to pass the Turing test?•Natural language processing: to communicate with examiner.•Knowledge representation: to store and retrieve information provided before or during interrogation.•Automated reasoning: to use the stored information to answer questions and to draw new conclusions.•Machine learning: to adapt to new circumstances and to detect and extrapolate patterns.•Vision (for Total Turing test): to recognize the examiner’s actions and various objects presented by the examiner.•Motor control (total test): to act upon objects as requested.•Other senses (total test): such as audition, smell, touch, etc.CS 561, Session 284What would a computer need to pass the Turing test?•Natural language processing: to communicate with examiner.•Knowledge representation: to store and retrieve information provided before or during interrogation.•Automated reasoning: to use the stored information to answer questions and to draw new conclusions.•Machine learning: to adapt to new circumstances and to detect and extrapolate patterns.•Vision (for Total Turing test): to recognize the examiner’s actions and various objects presented by the examiner.•Motor control (total test): to act upon objects as requested.•Other senses (total test): such as audition, smell, touch, etc.Core of the problem,Main focus of 561CS 561, Session 285What is an (Intelligent) Agent?•Anything that can be viewed as perceiving its environment through sensors and acting upon that environment through its effectors to maximize progress towards its goals.•PAGE (Percepts, Actions, Goals, Environment)•Task-specific & specialized: well-defined goals and environmentCS 561, Session 286Environment typesEnvironment AccessibleDeterministicEpisodic Static DiscreteOperating SystemVirtual RealityOffice EnvironmentMarsCS 561, Session 287Environment typesEnvironment AccessibleDeterministicEpisodic Static DiscreteOperating SystemYes Yes No No YesVirtual RealityYes Yes Yes/No No Yes/NoOffice EnvironmentNo No No No NoMars No Semi No Semi NoThe environment types largely determine the agent design.CS 561, Session 288Agent types•Reflex agents•Reflex agents with internal states•Goal-based agents•Utility-based agentsCS 561, Session 289Reflex agentsCS 561, Session 2810Reflex agents w/ stateCS 561, Session 2811Goal-based agentsCS 561, Session 2812Utility-based agentsCS 561, Session 2813How can we design & implement agents?•Need to study knowledge representation and reasoning algorithms•Getting started with simple cases: search, game playingCS 561,
View Full Document
Unlocking...