agent is anything that can be viewed as perceiving its environment through sensors and acting upon that environment through actuators. Human agent: eyes, ears, and other organs for sensors; hands, legs, mouth, and other body parts for actuators. Robotic agent: cameras and infrared range finders for sensors, various motors for actuators. A software agent receives keystrokes, file contents, and network packets as sensory inputs and acts on the environment by displaying on the screen, writing files, and sending network packets. Percept refers to the agent’s perceptual inputs at any given instant. An agent’s percept sequence is the complete history of everything the agent has ever perceived. Agent’s behavior is described by the agent function which maps from percept histories to actions: [f: P* à A].One way is to maintain an internal state(e.g. Cellphone knows its battery usage).Internal state can contain information about the state of the external environment.The state depends on the history of percepts and on the history of actions taken:[f: P*, A* S A] where S is the set of states Agent program runs on the physical architecture to produce f. Agent function is usually a table which is an external characterization of the agent. Internally, the agent function for an artificial agent will be implemented by an agent program. It is important to keep these two ideas distinct. The agent function is an abstract mathematical description; the agent program is a concrete implementation, running within some physical system. Vacuum cleaner example: Percepts: location and contents, e.g., [A,Dirty] Actions: Left, Right, Suck, NoOp. Agent’s function is look-up table. Rationality: Performance measuring success, Agents prior knowledge of environment, Actions that agent can perform, Agent’s percept sequence to date. Rational Agent: For each possible percept sequence, a rational agent should select an action that is expected to maximize its performance measure, given the evidence provided by the percept sequence, and whatever built-in knowledge the agent has. Rational is different from omniscience. An omniscient agent knows the actual outcome of its actions and can act accordingly; but this is impossible in reality. Percepts may not supply all relevant information E.g., in card game, don’t know cards of others. Rational is different from being perfect. Rationality maximizes expected outcome while perfection maximizes actual outcome.We consider the consequences of the agent’s behavior. Sequence of actions generated by an agent according to the percepts it receives causes the environment to go through a sequence of states. If the sequence is desirable, then the agent has performed well. This notion of desirability is captured by a performance measure that evaluates any given sequence of environment states. Autonomy of an agent is the extent to which its behaviour is determined by its own experience, rather than knowledge of designer. Extremes: No autonomy – ignores environment/data. Complete autonomy – must act randomly/no program. Example: baby learning to crawl. Ideal: design agents to have some autonomy and then have them possibly become more autonomous with experience.PEAS: Performance measure, Environment, Actuators, Sensors. Specifies the setting for designing an intelligent agent.Agent: Part-picking robot.Performance measure: Percentage of parts in correct bins.Environment: Conveyor belt with parts, bins.Actuators: Jointed arm and hand.Sensors: Camera, joint angle sensors.Agent: Interactive Spanish tutor.Performance measure: Maximize student’s score on test.Environment:Set of students.Actuators: Screen display (exercises, suggestions, corrections) Sensors: Keyboard.Self-driving car.Performance measure: Safe, fast, legal, comfortable trip, maximize profits, minimize accidents.Environment: Roads, other traffic, pedestrians, customers. Actuators: Steering wheel, accelerator, brake, signal, horn.Sensors: Cameras, sonar, speedometer, GPS, odometer, engine sensors, keyboard.Fully observable: everything agent requires to choose its actions is available to it via its sensors. If agent must make informed guesses about world, partially observable.Deterministic: next state of the environment is completely determined by the current state and the action executed by the agent. If aspects beyond the control of the agent and utility functions have to guess at changes in world then stochastic. Episodic: choice of current action isn’t dependent on previous actions. Sequential: current choice will affect future actions. Static: environment doesn’t change while agent is deciding over what to do. Dynamic: environments change so agent should/could console the world when choosing actions. Semidynamic: environment itself doesn’t change with time but agent’s score does.discrete: limited number of clear actions vs continuous: range of values. Single agent: agent operating by itself in environment vs multiagent: many agents working together. Known :o utcomes for all actions are given vs unknown:agent has to learn how environment works to make good decisions. Crossword: fully,deterministic,sequential,static, discrete,single.Poker:partially,stochastic,sequential,static,discrete,multi.Taxidriver:partially,stochastic,sequential,dynamic,conti,multi.partrobot:partially,stochastic,episodic,dynamic,conti,single.Imageanalysis:fully,deterministic,episodic,semi,conti,single.Four agent types: simple reflex agents, reflex agents with state/model, goal-based agents, utility-based agents. All of these agents can be turned into learning agents.Simple reflex agent:simple but very limited intelligence.Action doesn’t depend on percept history, only on current percept.Ex:Thermostat.
Therefore no memory requirements and could be infinite loops.Suppose vacuum cleaner does not observe location. What do you do given location = clean? Left on A or right on B -> infinite loop.Fly buzzing around window or light.Possible Solution: Randomize action.