0% found this document useful (0 votes)
12 views6 pages

Intelligent Agents in AI Lecture 2

The lecture covers the concept of intelligent agents in artificial intelligence, detailing their perception, actions, and rationality. It introduces the PEAS framework for task environments and discusses different types of agents, including simple reflex, model-based, goal-based, and utility-based agents. The importance of rationality in agent performance and the relationship between agents and their environments is emphasized.

Uploaded by

Doğukan Duman
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
12 views6 pages

Intelligent Agents in AI Lecture 2

The lecture covers the concept of intelligent agents in artificial intelligence, detailing their perception, actions, and rationality. It introduces the PEAS framework for task environments and discusses different types of agents, including simple reflex, model-based, goal-based, and utility-based agents. The importance of rationality in agent performance and the relationship between agents and their environments is emphasized.

Uploaded by

Doğukan Duman
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

BLG521E - Lecture 2

Outline BLG521E AdvancedArtificialIntelligence


Lecture2:IntelligentAgents
AgentsandEnvironments Rationality PEAS EnvironmentTypes AgentTypes

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

Agents

Agents
Agentsperceivetheirownactions
Effects?

Percept:theagentperceptualinput Perceptsequence:thecompletehistory Actionchoicesdependontheperceptsequence A i h i d d h Agentfunction,abstractmathematicaldescription (agentsbehavior) Agentprogramimplementsthefunction

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

VacuumCleanerWorld

AVacuumCleanerAgentFunction

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

BLG521E - Lecture 2

AVacuumCleanerAgentFunction

VacuumCleanerAgent PM
Theamountofdirtcleanedupinasingleeighthour shift. Rewardingagentforhavingacleanfloor. Factoringamountofelectricityconsumedandthe amountofnoisegenerated amount of noise generated DesignPMaccordingtowhatyouwant

What is the right way to fill out the table?

What makes an agent good, bad, intelligent or stupid?


AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link] 7 AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link] 8

Rationality
Arationalagentdoestherightthing Whatisrationalatanygiventimedependson:
Theperformancemeasurethatdefinesthecriterionof success The agents prior knowledge of the environment Theagent spriorknowledgeoftheenvironment Theactionsthattheagentcanperform Theagentsperceptsequencetodate

RationalAgent
Foreachpossibleperceptsequence,arationalagent shouldselectanactionthatisexpected tomaximize itsperformance efficiency,giventheevidence providedbytheperceptsequenceandwhatever builtinknowledgetheagenthas. g g Witharationalagent,whatyouaskiswhatyouget

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

10

[Link]
Omniscienceisimpossibleinreality Agentsdontestimatetheactualoutcomeofactions Rationalitymaximizesexpectedoutcome,while perfectionmaximizesactualperformance f ti i i t l f

FurtherDimensionsinRationality
Informationgathering
Exploration Helpsmaximizetheexpectedoutcome

Learning Autonomy Withorwithoutinitialknowledge

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

11

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

12

BLG521E - Lecture 2

Agents<>Environments
Taskenvironmentformstheproblem
Rationalagentsarethesolutions

TheNatureofEnvironments
PEASfortaskenvironments:
Performancemeasure Environment Actuators Sensors

Thetaskenvironmentaffectstheappropriatedesign oftheagent of the agent

PEASforautomatedtaxidriver

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

13

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

14

PropertiesofTaskEnvironments
[Link] [Link]
strategic

ExamplesofTaskEnvironments
Task Environment Crossword puzzle Chesswitha clock Poker Backgammon Taxidriving Observable Deterministic Episodic Static Discrete Agents

[Link] S i [Link] D i
semidynamic

[Link] [Link]
competitive cooperative
AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link] 15

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

16

ExamplesofTaskEnvironments
Task Environment Crossword puzzle Chesswitha clock Poker Backgammon Taxidriving Observable Deterministic Episodic Static Discrete Agents

ExamplesofTaskEnvironments
Task Environment Crossword puzzle Chesswitha clock Poker Backgammon Taxidriving Observable Deterministic Episodic Static Discrete Agents

Fully

Deterministic

Sequential

Static

Discrete

Single

Fully

Deterministic

Sequential

Static

Discrete

Single

Fully

Strategic

Sequential

Semi

Discrete

Multi

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

17

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

18

BLG521E - Lecture 2

ExamplesofTaskEnvironments
Task Environment Crossword puzzle Chesswitha clock Poker Backgammon Taxidriving Observable Deterministic Episodic Static Discrete Agents

ExamplesofTaskEnvironments
Task Environment Crossword puzzle Chesswitha clock Poker Backgammon Taxidriving Observable Deterministic Episodic Static Discrete Agents

Fully

Deterministic

Sequential

Static

Discrete

Single

Fully

Deterministic

Sequential

Static

Discrete

Single

Fully

Strategic

Sequential

Semi

Discrete

Multi

Fully

Strategic

Sequential

Semi

Discrete

Multi

Partially

Stochastic

Sequential

Static

Discrete

Multi

Partially Fully

Stochastic Stochastic

Sequential Sequential

Static Static

Discrete Discrete

Multi Multi

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

19

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

20

ExamplesofTaskEnvironments
Task Environment Crossword puzzle Chesswitha clock Poker Backgammon Taxidriving Observable Deterministic Episodic Static Discrete Agents

TheStructureofAgents
ThejobofAIistodesigntheagentprogram

Fully

Deterministic

Sequential

Static

Discrete

Single

Agentarchitecture Agent=Architecture+Program

Fully

Strategic

Sequential

Semi

Discrete

Multi

Partially Fully Partially

Stochastic Stochastic Stochastic

Sequential Sequential Sequential

Static Static Dynamic

Discrete Discrete Continuous

Multi Multi Multi

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

21

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

22

AgentTypes
Simplereflexagents Modelbasedreflexagents Goalbasedagents Utilitybasedagents

SimpleReflexAgents

Alltheseagentscanbeconvertedintolearning agents

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

23

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

24

BLG521E - Lecture 2

ReflexVacuumAgentProgram

SimpleReflexAgentProgram

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

25

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

26

ModelbasedReflexAgents

ModelbasedAgentProgram

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

27

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

28

GoalBasedAgents

UtilitybasedAgents

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

29

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

30

BLG521E - Lecture 2

LearningAgents

AdvancedArtificialIntelligence(BLG521E)@ITU::ComputerEngineeringDepartment,[Link]

31

You might also like