Today’s Class Artificial Intelligence Class 2: Intelligent Agents • What’s an agent? • Definition of an agent • Rationality and autonomy • Types of agents • Properties of environments Dr. Cynthia Matuszek – CMSC 671 3 Pre-Reading: Quiz What is an Agent? • What are sensors and percepts? • An intelligent agent is: • A (usually) autonomous entity which… • Observes an environment (the world) • What are actuators (aka effectors) and actions? Shows � • Acts on its environment in order to achieve goals “agency” • An intelligent agent may learn • What are the six environment characteristics that R&N use to characterize different problem spaces? • Not always • A simple “reflex agent” still counts as an agent Observable Deterministic Static • Behaves in a rational manner # of Agents Episodic Discrete • Not “optimal” 4 5 Human Sensors/Percepts, How Do You Design an Agent? Actuators/Actions • An intelligent agent: • Sensors: • Eyes (vision), ears (hearing), skin (touch), tongue (gustation), nose • Perceives its environment via sensors (olfaction), neuromuscular system (proprioception), … • Acts upon that environment with its actuators (or • Percepts: “that which is perceived” effectors ) • At the lowest level – electrical signals from these sensors • Properties: • After preprocessing – objects in the visual field (location, textures, colors, …), auditory streams (pitch, loudness, direction), … • Autonomous • Reactive to the • Actuators/effectors: environment • Limbs, digits, eyes, tongue, … • Pro-active (goal-directed) • Actions: • Interacts with other agents via the environment • Lift a finger, turn left, walk, run, carry an object, … 6 7 1
Human Sensors/Percepts, E.g.: Automated Taxi Actuators/Actions • Sensors: • Percepts: Video, sonar, speedometer, odometer, engine • Eyes (vision), ears (hearing), skin (touch), tongue (gustation), nose sensors, keyboard input, microphone, GPS, … (olfaction), neuromuscular system (proprioception), … • Actions: Turn, accelerate, brake, speak, display, … • Percepts: “that which is perceived” The Point: • Goals: Maintain safety, reach destination, maximize • At the lowest level – electrical signals from these sensors • Percepts and actions need profits (fuel, tire wear), obey laws, provide passenger • After preprocessing – objects in the visual field (location, textures, colors, to be carefully defined …), auditory streams (pitch, loudness, direction), … comfort, … • Sometimes at different • Actuators/effectors: • Environment: U.S. urban streets, freeways, traffic, levels of abstraction! • Limbs, digits, eyes, tongue, … pedestrians, weather, customers, … • Actions: Different aspects of driving may require � • Lift a finger, turn left, walk, run, carry an object, … different types of agent programs. 8 9 Rationality PEAS • An ideal rational agent , in every possible world state, does • Agents must have: action(s) that maximize its expected performance • P erformance measure • Based on: • The percept sequence (world state) • E nvironment • Its knowledge (built-in and acquired) • A ctuators • Rationality includes information gathering • If you don’t know something, find out! • S ensors • No “rational ignorance” • Need a performance measure • Must first specify the setting for intelligent agent • False alarm (false positive) and false dismissal (false negative) rates, design speed, resources required, effect on environment, constraints met, user satisfaction, … 10 PEAS PEAS • Agent: Part-picking robot • Agent: Interactive English tutor • Performance measure: Percentage of parts in • Performance measure: Maximize student's score on correct bins test • Environment: Conveyor belt with parts, bins • Environment: Set of students • Actuators: Jointed arm and hand • Actuators: Screen display (exercises, suggestions, corrections) • Sensors: Camera, joint angle sensors • Sensors: Keyboard 2
PEAS: Setting PEAS • Agent: Medical diagnosis system • Specifying the setting • Consider designing an automated taxi driver: • Performance measure: Healthy patient, minimize costs, • Performance measure? Safe, fast, legal, comfortable trip, lawsuits maximize profits • Environment: Patient, hospital, staff • Environment? Roads, other traffic, pedestrians, customers • Actuators: Screen display (questions, tests, diagnoses, • Actuators? Steering wheel, accelerator, brake, signal, horn treatments, referrals) • Sensors? Cameras, sonar, speedometer, GPS, odometer, • Sensors: Keyboard (entry of symptoms, findings, engine sensors, keyboard patient's answers) Autonomy Some Types of Agent • An autonomous system is one that: 1. Table-driven agents • Determines its own behavior • Use a percept sequence/action table to find the next action • Implemented by a (large) lookup table • Not all its decisions are included in its design 2. Simple reflex agents • It is not autonomous if all decisions are made by its • Based on condition-action rules designer according to a priori decisions • Implemented with a production system • “Good” autonomous agents need: • Stateless devices which do not have memory of past world states • Enough built-in knowledge to survive 3. Agents with memory • The ability to learn • Have internal state • Used to keep track of past states of the world • In practice this can be a bit slippery 18 19 Some Types of Agent (1) Table-Driven Agents • Table lookup of: 4. Agents with goals • Percept-action pairs mapping • Have internal state information, plus… • Every possible perceived state ßà optimal • Goal information about desirable situations action for that state • Agents of this kind can take future events into consideration • Problems: 5. Utility-based agents • Too big to generate and store • Chess has about 10 120 states, for example • Base their decisions on classic axiomatic utility theory • Don’t know non-perceptual parts of state • In order to act rationally • E.g., background knowledge • Not adaptive to changes in the environment • Must update entire table • No looping • Can’t condition actions on previous actions/states 20 www.quora.com/How-do-you-know-if-your-chess-pieces-are-in-strategic-positions 3
(2) Simple Reflex Agents (1) Table-Driven/Reflex Agent • Rule-based reasoning • To map from percepts to optimal action • Each rule handles a collection of perceived states • “If your rook is threatened…” • Problems • Still usually too big to generate and to store • Still no knowledge of non-perceptual parts of state • Still not adaptive to changes in the environment • Change by updating collection of rules • Actions still not conditional on previous state 22 (3) Architecture for an (3) Agents With Memory Agent with Memory • Encode “internal state” of the world • Used to remember the past (earlier percepts) • Why? • Sensors rarely give the whole state of the world at each input • So, must build up environment model over time • “State” is used to encode different “world states” • Different worlds generate the same (immediate) percepts • Requires ability to represent change in the world • Could represent just the latest state • But then can’t reason about hypothetical courses of action 24 25 (4) Architecture for (4) Goal-Based Agents Goal-Based Agent • Choose actions that achieve a goal • Which may be given, or computed by the agent • A goal is a description of a desirable state • Need goals to decide what situations are “good” • Keeping track of the current state is often not enough • Deliberative instead of reactive • Must consider sequences of actions to get to goal • Involves thinking about the future • “What will happen if I do...?” 27 28 4
Recommend
More recommend
