AI 2 - Intelligent Agents

percept sequence

A ________ ________ is the complete history of everything the agent has ever percieved

evolves independently, actions, world

A model based reflex agent requires two types of knowledge to be modeled: -Information about how the world _________ _________ -information about how the agent's _______ affect the ________

competitive multiagent

A multi-agent environment in which actions of one agent affects the actions of other agents

cooperative multiagent

A multi-agent environment in which agents work together for a common cause

Agent function

Abstract mathematical concepts that describe an agent

actuators

An agent acts upon the environment using ________

behavior, maps, action

An agent function describes the agent's ________ and ________ any given percept sequence to a(n) _________

sensors

An agent perceives its environment using ________

goal based agent

An agent that asks "what will happen if I do this" and "will that action make me happy?" - need some kind of information that describes desireable situations

simple reflex agent

An agent that selects actions based on the current percept and ignores the percept history

model based reflex agent

An agent that tracks part of the world that it can't see. It maintains an internal state that depends on percept history and reflects some unobserved aspects of the current state

utility based agent

An agent that uses an internal function that agrees with an external performance measure to maximize its usefullness.

partially observable environment

An environment in which an agent's sensors do not give it access to the complete state of the environment at each point in time

fully observable environment

An environment in which an agent's sensors give it access to the complete state of the environment at each point in time

Single Agent environment

An environment in which one agent operates alone in the environment

Multiple Agent environment

An environment in which several agents operate in the environment

known environment

An environment in which the agent knows all the outcomes for all actions or their probabilities - requires no learning

unknown environment

An environment in which the agent must learn how the environment works in order to make good decisions

episodic environment

An environment in which the agent perceives the environment and executes a single action, the choice of each action only depends on the current state.

sequential environment

An environment in which the current decision could affect all future decisions

deterministic environment

An environment in which the next state is completely determined by the current state

stochastic environment

An environment in which the next state is not completely determined by the current state

dynamic environment

An environment that can change while the agent is deliberating

semidynamic environment

An environment that does not change but the agent's performance score does change over time

static environment

An environment that does not change while the agent is deliberating

discrete environment

An environment that has a limited number of distinct, clearly defined percepts and actions. Applies to the way time is handled

continuous environment

An environment that has no clear division of percepts and actions along time. Applies to the way time is handled

Agent program

Concrete implementation of an agent

table, percept sequences, behaviors

In external characterization, can create a ______ given an agent, that describes the agent based on its _________ _________ and __________

agent program, agent function

In internal characterization, the ______ _______ is the internal implementation of the _______ _______

percept

The agent's perceptual inputs at any given instant

Environment

The problem space

Performance Measure

What the agent wants to maximize