Agents and State Spaces

Question 1

Rational Agent

Answer 1

Chooses action that maximizes the expected performance measure given sequence of observed perceptions

Question 1

Performance Measure

Answer 1

Evaluates sequence of configurations of environment, chosen by creator

Question 2

Unknown VS Known Environment

Answer 2

Whether environment is known beforehand

Question 3

Types of Known Environments

Answer 3

1. Observable vs Fully/Partially
2. Single-Agent vs Multi
3. Deterministic vs Stochastic
4. Episodic vs Sequential
5. Static vs Dynamic
6. Discrete vs Continuous

Question 4

Observable vs Fully/Partially Environment

Answer 4

Can agent sense all relevant aspects of the environment directly?

Question 5

Single vs Multi Agent Environment

Answer 5

More than one agent acting at same time? Based on intention

Question 6

Deterministic vs Stochastic Environment

Answer 6

Do actions always have same outcome?

Question 7

Static vs Dynamic Environment

Answer 7

Can environment change while agent is thinking?

Question 8

Discrete vs Continuous Environment

Answer 8

Is there a finite number of possible ways to get to the goal?

Question 9

Agents Architectures

Answer 9

1. Table Driven Agent
2. Simple Reflex Agent
3. Model-Based Reflex Agent
4. Goal-Based Agent
5. Utility-Based Agent
6. Learning Agent

Question 10

Table Driven Agent

Answer 10

Fills in table that holds all possible actions using perception history as key, can use table for future actions, inefficient

Question 11

Simple Reflex Agent

Answer 11

Uses table of rules and current state as key to lookup action

Question 12

State

Answer 12

Encoding of current configuration of the environment

Question 13

Model-Based Reflex Agent

Answer 13

Handles partial-observability.
Uses sensors, state, and known physics of the world as a key in rule table to lookup action

Question 14

Goal-Based Agent

Answer 14

Handles sequential environments.
Plans one step into the future.
Uses sensors, state, and known physics of the world to predict future states which are used as a key in GOAL table to lookup action

Question 15

Utility-Based Agent

Answer 15

Handles stochastic environments.
Uses probability to find expected ‘goodness’ of future states.
Uses sensors, state, and known physics of the world to predict future states which are used as a key in UTILITY table to lookup action

Question 16

Learning Agent

Answer 16

Handles unknown environments.
Sensors give info to performance element that then uses the critic, learning element, and problem generator to determine action

Question 17

Performance Element

Answer 17

Evaluates states/actions, provides info about environment

Question 18

Learning Element

Answer 18

Improves internal model

Question 19

Problem Generator

Answer 19

Suggests actions to improve learning

Question 20

Critic

Answer 20

Evaluates overall performance of agent

Question 21

‘Physics’ of Environment

Answer 21

Dictate how state changes over time in response to agent actions

Question 22

State Space

Answer 22

Set of all possible states

Question 23

State Space - Graph Form

Answer 23

For problems with finite number of states, can represent entire state space and transitions with a graph