Transcript agent
ARTIFICIAL INTELLIGENCE [INTELLIGENT AGENTS PARADIGM] SIMPLE INTELLIGENT AGENTS Professor Janis Grundspenkis Riga Technical University Faculty of Computer Science and Information Technology Institute of Applied Computer Systems Department of Systems Theory and Design E-mail: [email protected] Intelligent Agents and Their Programs • An agent is just something that perceives and acts • Variety of definitions • Agent functions: mapping percepts to actions Rational Agent • One that does the right thing • How and when to evaluate the agents success? • What is rational depends on four factors: – Performance measure (for the how?) – Percept sequence – Knowledge about the environment – Actions • Ideal rational and omniscient agents Autonomous Agent • One whose actions are not based completely on built-in knowledge • One whose actions are based on both built-in knowledge and own experience • Initial knowledge provides an ability to learn • A truly autonomous agent can adapt to a wide variety of environments Structures of Intelligent Agents • Agent is a program and an architecture • Initial phase for agent program is to understand and describe: – Percepts – Actions – Goals – Environment Agent Programs (1) • Skeleton-Agent > Single percept Update-Memory(memory, percept) Choose-Best-Action(memory) > Action Update-Memory(memory, action) Return: action Agent Programs (2) • Table-Driven-Agent > Percept sequences Look-Up(percepts, table) Return: action Agent Programs (3) • Rule-Based Agent (simple reflex agent) > Percept Interpretation(percept) > Rule match Interpreted percept IF pattern > Rule Firing THEN pattern > Action Return: action Applications: logical reasoning systems Agent Programs (4) • Model-Based Agent > Percept Update-State(state, percept) > Rule match State IF pattern > Rule Firing THEN pattern > Action Return: action Applications: logical reasoning systems, decision making agents Agent Programs (5) • Goal-Based Agent > Goal > Inference > Search and Planning Applications: planning agents Agent Programs (6) • Utility-Based Agent > Utility Applications: game playing, decision making agents