Comparing reflex,

goal-based, and
utility-based
agents in AI.
Presented by – Souvik Nayak
STD – BWU/MCA/23/123
Course name – Artificial Intelligence
and Machine Learning
Course Code – MCA301
TABLE OF CONTENTS

Introduction & Reflex Agents Goal-Based

classification Agents

Utility-Based Comparison of Applications &

Agents Agent Types Conclusion
INTRODUCTIO
N
An AI agent is a system that can perceive its
environment, think intelligently, and take
actions to achieve specific goals. It works
autonomously, using techniques like machine
learning and natural language processing, to
solve problems and interact with users or
systems.
Examples: Virtual assistants (Siri, Alexa),Self-
driving cars, Customer service chatbots
 Classification

2. Goal-based
1. Reflex agents 3. Utility-based
agents agents
Make decisions by
Respond directly to evaluating actions
that help achieve Consider both goals
stimuli using simple
specified goals. and the overall utility
condition-action rules
(or preference) of
without considering
outcomes to make
the environment's
optimal decisions.
history.
 Reflex Agents
 Definition : Operate based on condition-action rules (e.g., "if-then" statements).
Do not consider the history or future consequences of actions.

Real-World
Key Features Advantages Limitations Examples
Simple and reactive. Efficient in Cannot handle Thermostat adjusting
predictable complex or dynamic temperature based on
environments. environments. room conditions.

Quick decision- Easy to design and Lack memory or Traffic light systems
making based on the implement. planning capability. responding to sensor
current state. inputs.
 Goal-Based Agents
 Definition : Make decisions by considering how actions help achieve predefined goals.
Use search and planning algorithms to evaluate future states.

01 02 03 04
Real-World
Key Features Advantages Limitations
Examples
 Require well-  GPS navigation
 Actions are guided  More flexible than
defined goals. systems calculating
by a specific goal. reflex agents. 
  Computationally routes to a
Can incorporate Capable of handling
more expensive destination.
history and predict dynamic and
than reflex agents.  Chess-playing AI bots
future states. partially observable
environments. aiming to checkmate
opponents.
  Utility-Based Agents
 Definition : Extend goal-based agents by evaluating utility (desirability) of different states.
Use a utility function to make trade-offs and optimize actions.

Real-World
Key Features Advantages Limitations Examples
● Actions aim to ● Highly adaptive ● Require well- ● Self-driving cars
maximize overall utility. and flexible. designed utility balancing safety,
● Handle multiple ● Provide optimal functions. speed, and fuel
objectives or conflicting solutions efficiency.
● Computationally ● Investment advisors
goals. considering trade- complex due to optimizing portfolios
offs. utility evaluation. based on risk and
return.
 Comparison of Agent Types
Aspect Reflex Agents Goal-Based Utility-Based
Agents Agents
Decision Basis Current Achievement of Maximization of
environment state predefined goals utility
Flexibility Low Medium High

Complexity Simple Moderate High

Adaptability Rigid Flexible Highly flexible

Examples Thermostat, traffic GPS navigation, Self-driving cars,

lights. chess-playing investment
bots. advisors.
 Applications

Reflex Goal-Based Agents: Utility-Based

Agents: Agents:
Suitable for Used in applications Ideal for complex,
predictable, real-time requiring planning and multi-objective
systems (e.g., strategic decisions systems (e.g.,
robotics, sensors). (e.g., game AI, autonomous vehicles,
navigation). smart cities).
 Conclusion
● Key Insights: The choice of an agent depends
on the complexity of the task and the
environment.
● Advancements in AI: Modern agents often
combine these approaches for hybrid
decision-making.
● Future Scope: Enhanced AI agents to solve
real-world problems with even higher
efficiency and adaptability.
 references
 Artificial Intelligence: A Modern Approach by Stuart Russell and Peter
Norvig.
 https://www.geeksforgeeks.org/types-of-agents-in-ai/
 https://www.javatpoint.com/types-of-ai-agents
 https://www.geeksforgeeks.org/agents-artificial-intelligence/
THANKS!
Do you have any
questions?

Credits: This presentation template was

created by Slidesgo, and includes icons
by Flaticon and infographics & images
by Freepik
Please keep this slide for attribution