Artificial Intelligence Course of BIT Unit 2

Introduction of agents, Structure(configuration) of Intelligent agent,
Properties of Intelligent Agents
2.2. PEAS Description of Agents
2.3. Types of Agents: Simple Reflexive, Model Based, Goal Based, Utility Based,
Learning agent.
2.4. Types of Environments: Deterministic/Stochastic, Static/Dynamic,
Observable/Semi-observable, Single Agent/Multi Agent

1. Unit 2
Intelligent Agents
2.1. Introduction of agents, Structure(configuration) of Intelligent agent,
Properties of Intelligent Agents
2.2. PEAS description of Agents
2.3. Types of Agents: Simple Reflexive, Model Based, Goal Based, Utility Based,
Learning agent.
2.4. Types of Environments: Deterministic/Stochastic, Static/Dynamic,
Observable/Semi-observable, Single Agent/Multi Agent
Prepared by Shiv Raj Pant1

2. What is an agent ?
•
An Intelligent Agent perceives its environment via sensors and acts rationally upon that
environment with its effectors (actuators).
•
Hence, an agent gets percepts one at a time, and maps this percept sequence to actions .
Properties of the agent
–Autonomous
–Interacts with other agents plus the environment
–Reactive to the environment
–Pro-active (goal-directed)
Prepared by : Shiv Raj Pant
2

3. sensors/percepts and effectors/actions
For Humans
–Sensors: Eyes (vision), ears (hearing), skin (touch), tongue (gestation), nose (olfaction),
neuromuscular system (proprioception)
–Percepts: At the lowest level –electrical signals from these sensors. After preprocessing:
objects in the visual field (location, textures, colors, …), auditory streams
(pitch, loudness, direction), …
–Effectors: limbs, digits, eyes, tongue, …..
–Actions: lift a finger, turn left, walk, run, carry an object, …
Automated taxi driving system
-sensors: Video camera, speedometer, odometer, engine sensors, keyboard input, microphone,
GPS, …
-Actions: Steer, accelerate, brake, horn, speak/display, …
-Performance measures: Maintain safety, reach destination, maximize profits (fuel, tire wear),
obey laws, provide passenger comfort, …
-Environment: Urban streets, freeways, traffic, pedestrians, weather, customers, …
Prepared by : Shiv Raj Pant
3

4. •The agent function is mathematical concept that maps percept sequence to actions.
•The agent function will internally be represented by the agent program.
•The agent program is concrete implementation of agent function it runs on the physical
architecture to produce f.
Prepared by : Shiv Raj Pant
4

5. The vacuum-cleaner world: Example of Agent
•Consider an automated (intelligent) vacuum
cleaner machine.
•Suppose there are two rooms A and B. The job
of the vacuum cleaner is to sense the dirt
in the room and clean the rooms. The vacuum
cleaner can move from one room to another
(left or right)
Environment: square A and B
Percepts: [location, content] E.g. [A, Dirty]
Actions: left, right, clean, and no-op
Prepared by : Shiv Raj Pant
5
Percept sequenceAction
[A,Clean] Right
[A, Dirty] Suck
[B, Clean] Left
[B, Dirty] Suck
………. ……
The agent can be made to work by implementing
a simple condition-action table as shown
below.
Every time the agent senses the environment
and matches the percept with the condition in
the table. It then executes the defined task.

6. Rationality of an intelligent agent
•A rational agent is one that does the right thing. i.e. Every entry in the table is filled
out correctly.
What is the right thing?
–Right action is the one that will cause the agent to be most successful.
•Therefore we need some way to measure success of an agent.
•Performance measures are the criterion for success of an agent behavior.
E.g., performance measure of a vacuum-cleaner agent could be amount of dirt cleaned up, amount
of time taken, amount of electricity consumed, amount of noise generated, etc.
It is better to design Performance measure according to what is wanted in the environment
instead of how the agents should behave.
It is not easy task to choose the performance measure of an agent. For example if the
performance measure for automated vacuum cleaner is “The amount of dirt cleaned within a
certain time” Then a rational agent can maximize this performance by cleaning up the dirt ,
then dumping it all on the floor, then cleaning it up again , and so on. Therefore “How clean
the floor is” is better choice for performance measure of vacuum cleaner.
Prepared by : Shiv Raj Pant
6

7. Therefore, What is rational at a given time depends on four things:
–Performance measure,
–Prior environment knowledge,
–Actions,
–Percept sequence to date (sensors).
Definition: A rational agent chooses whichever action maximizes the expected value of the
performance measure given the percept sequence to date and prior environment knowledge.
Prepared by : Shiv Raj Pant
7

8. Environments
•To design a rational agent we must specify its task environment.
•Task environment means: PEAS description of the environment:
–Performance
–Environment
–Actuators
–Sensors
Example: Fully automated taxi:
PEAS description of the environment:
Performance: Safety, destination, profits, legality, comfort
Environment: Streets/freeways, other traffic, pedestrians, weather,, …
Actuators: Steering, accelerating, brake, horn, speaker/display,…
Sensors: Video, speedometer, engine sensors, keyboard, GPS, …
Prepared by : Shiv Raj Pant
8

9. Types of Agent
1. Simple Reflex Agent
•Store percept-action pairs defining all possible condition-action rules necessary to
interact in an environment. The agents acts by looking up in the table (table look-up).
Problems
–Too big to generate and to store (for example, Chess has about 10^120 states)
–No knowledge of non-perceptual parts of the current state (the agent can’t work in
uncertainity( if some sensors fail)
-The agent acts fully on current state. It can not remember past percepts/actions.
–Not adaptive to changes in the environment; requires entire table to be updated if changes
occur
Prepared by : Shiv Raj Pant
9

E
nv
ir
on
m
en
t
Agent
What the world
is like now
What action I
should do now
Condition-action rules
Sensors
Effectors

10. 2. Model-based Agent (Reflex Agent with Internal State)
•Encode "internal state" of the world to remember the past as contained in earlier percepts
•Sometimes, sensors do not give the entire state of the world at each input. In such partially
oberservableenvironment, a simple reflex agent does not work. So perception of the
environment is captured over time. "State" used to encode different "world states" that
generate the same immediate percept.
•Complex than simple reflex agent. Requires ability to represent change in the world
Prepared by : Shiv Raj Pant
10

11. 3. Goal-Based Agent
•Sometimes an agent is goal directed.
i.e. it looks for a goal to decide
which action to take
• A goal based agent choose actions so as
to achieve a (given or computed) goal.
• A goal is a description of a desirable
situation
• this type of agent is deliberative
instead of reactive
• the agent may have to consider long
sequences of possible actions before
deciding if goal is achieved --
involves consideration of the future,
“what will happen if I do...?”
Prepared by : Shiv Raj Pant
11

12. 4. Utility-Based Agent
• When there are multiple possible
alternatives, how to decide which
one is best?
• A goal specifies a crude distinction
between a happy and unhappy state,
but sometimes we need a more general
performance measure that describes
"degree of happiness"
• The utility function
U:States →Reals
indicates a measure of success or
happiness at a given state
• Allows decisions comparing choice
between conflicting goals, and
choice between likelihood of success
and importance of goal (if
achievement is uncertain)
•For example, this type of agent
implementation is suitable for
automated taxi.
Prepared by : Shiv Raj Pant
12

13. 5. Learning agent
Prepared by : Shiv Raj Pant
13

14. Types of environments
Fully observable/ Partially Observable
–If an agent's sensors give it access to the
complete state of the environment needed to
choose an action, the environment is
accessible.
–Such environments are convenient, since the
agent is freed from the task of keeping track
of the changes in the environment.
Deterministic/ Non-deterministic
–An environment is deterministic if the next
state of the environment is completely
determined by the current state of the
environment and the action of the agent.
–In an accessible and deterministic environment
the agent need not deal with uncertainty.
Episodic/ Nonepisodic.
–An episodic environment means that subsequent
episodes do not depend on what actions
occurred in previous episodes.
–Such environments do not require the agent to
plan ahead.
Prepared by : Shiv Raj Pant
14
Static/ Dynamic.
–A static environment does not change while the
agent is thinking.
–In a static environment the agent need not
worry about the passage of time while he is
thinking, nor does he have to observe the
world while he is thinking.
–In static environments the time it takes to
compute a good strategy does not matter.
Discrete/ Continuous.
–If the number of distinct percepts and actions
is limited the environment is discrete,
otherwise it is continuous.
Single Agent/ Multi-Agent
–If an environment does not contain other
rationally thinking, adversary agents, the
agent need not worry about strategic, game
theoretic aspects of the environment
–Most engineering environments are without
rational adversaries, whereas most social and
economic systems get their complexity from the
interactions of (more or less) rational
agents.

15. End of Unit 2
Thank you!
Prepared by Shiv Raj Pant15