Agents

Question 1

What is an Agent

Answer 1

Anything that can perceive their environment using sensors and act on it through actuators

Question 2

What are the three types of agents

Answer 2

• Human Agent
• Robotic Agent
• Software Agent

Question 3

How do you analyse an Agents performance

Answer 3

PEAS
- Performance
- Environment
- Actuators
- Sensors

Question 4

What are the 5 Agent classifications

Answer 4

• Simple Reflex Agent
• Model-Based Agent
• Goal-Based Agent
• Utility-Based Agent
• Learning Agent
  *An agent can be more than one type of agent

Question 5

Name the Environmental Classifications

Answer 5

-Fully/Partially Observable
-Deterministic/Stohastic
-Episodic/Sequential
-Static/Dynamic
-Discrete/Continuous

Question 6

Describe a simple Reflex Agent

Answer 6

The most basic Agent that reacts to the environment through preset instructions with no regard to past outcomes

Question 7

describe a Model-Based Agent

Answer 7

An Agent that reacts to the environment through preset instructions with some consideration of past outcomes

Question 8

Describe a Goal-Based Agent

Answer 8

Selects a method to achieve a goal in the most optimal way

Question 9

Describe a Utility Based Agent

Answer 9

Finds the most desirable goal to work towards through methods, Used if the end goal is unclear

Question 10

describe a Learning Agent

Answer 10

Compares the results of previous actions to work towards the best result available

Question 11

What is the difference in the fully or partially Observable environment

Answer 11

-Fully Observable - everything is visible to the AI eg. Board Games like chess or monopoly
-Partially Observable - Some things are hidden from the AI eg. Real Life or Board Games like Poker

Question 12

Difference in a Deterministic or Stochastic

Answer 12

-Deterministic - The effect is Guaranteed eg. Moving a piece in Chess
-Stochastic - the effect is unclear about the outcome of an action eg. Raising in poker

Question 13

Difference in Episodic and Sequential Enviroments

Answer 13

-Episodic - each action performed is independent of the last
-sequential - each action performed is linked to the previous action

Question 14

Difference between a Static and Dynamic environment

Answer 14

• Static - the AI assumes that noting is going to change whilst Calculating an action
• Dynamic - Changing are occurring even when the AI is calculating

Question 15

Difference between Discrete and Continuous Environments

Answer 15

• Discrete - the AI has a finite number of possible actions
• Continuous- the AI has infinite number of possible actions