Intelligent Agents

Question 1

What’s an “agent”?

Answer 1

An agent is anything that can be viewed as percieving its “environment” through “sensors” and acting upon that environment through “actuators”

Question 2

What’s an “agent function”?

Answer 2

Mathematically speaking, we say that an agent behaviour is described by the “agent function” that maps any given percept sequence to an action.

Question 3

What’s a “performance measure”?

Answer 3

The “performance measure” is the evaluation of any given environment states, and is a mean to understand a good from a bad behaviour of an “agent”.

Question 4

What’s a “rational agent”?

Answer 4

For each possible percept sequence, a “rational agent” should select an action that is expect to maximize its “performance measure”, given the evidence provided by the “percept sequence” and whatever built-in knowledge the agent has.

Question 5

What is the “task environment”?

Answer 5

The “task environment” is conceptually the problem to which the agent should be the solution. It is defined via P.E.A.S (Performance, Environment, Actuator, Sensors) descriptions.

Question 6

What is the general structure of an agent?

Answer 6

In general an agent is made up of 2 parts: “architecture” and “program”. The “program” is the implementation of the “agent function”, the “architecture” is the device with sensors and actuators in which the “program” runs.

Question 7

“Table-driven agents”

Answer 7

The “table-driven agents” are conceptually the most basic for of agents, they have a complete table of (percept sequence, action)