Scribd
Upload
19 views

Assignment 1c

Uploaded by

shezy3466
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
19 views

Assignment 1c

Uploaded by

shezy3466
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

ARTIFICIAL INTELLIGENCE

Assignment 1
Name: Abdul Rehman

Roll no: 23-Arid-3023

Instructor: Ms. Mishal Fatima

TYPES OF AI AGENTS
1. Simple Reflex Agents
Meaning and Process:
A Simple Reflex Agent: What Is It?
An agent that selects actions based purely on current situations without consideration of any
history uses pre-specified condition-action rules. It is called a simple reflex agent.
How Does It Arrive at Decisions?
It bases its decisions on condition-action rules, or "if-then" logic. Having scanned the
environment, it applies the relevant rule and acts accordingly.
Real-Life Examples:

• Thermostats: Adjust heating based on outside temperature.


• Automatic doors: open because of sensors that sense movement.
• Light sensors: Switch on and off depending on the light intensity in the surrounding
space.

Limitation of Dynamic Environments:

• No learning memory: Do not learn or adapt to past events.


• Semi adaptability: only deals with already established circumstances.
• Not suitable for complicated goals: Cannot plan or deduce.
2. Model-Based Reflex Agents
Definition and Methodology:

A Model-Based Reflex Agent monitors nonvisible changes in the environment and takes
decisions based on the information available and also based on internal model of the world.

How is it Different from the Basic Reflex Agent?

Model-Based agents have an internal state and memory of previous states to make better
decision-making.

The World Model and the Internal State:

• Internal state: It monitors the sub-perceptible environmental parameters.


• World model: It assists in making decisions and predicting the consequence.

2. Use Cases
Practical Scenarios:

• Robots: Improve the process of navigation as it monitors the previous blockages.


• Autonomous vehicles: Predict the traffic trend by processing the data from past dates.

Why So Effective?

It becomes more elastic and brighter due to its ability to predict the changes as well as the
memory of the previous states in variable surroundings.

3. Goal-Based Agents
1. Definition and Operation: What Are Agents with a Goal?

• Goal-Based Agents make judgments based on future outcome-based action


determination to achieve certain goals, including long-term objectives.
• They evaluate the appropriateness of activities based on their capacity to help achieve
goals, enabling planning and a forward-thinking attitude.
2. Equivalence

• Simple Reflex Agents: They do not have any goals. They simply react to the situation at
hand.
• Model-Based Reflex Agents: These are concerned about the current state and have
some inner model.
• Goal-Based Agents: These agents can relate the short-term level of goals with a long-
term level, making them much more flexible in difficult situations. Goals are important
as they guide the agent towards achievement.

4. Utility-Based Agents
Definition and Working:

What is a Utility-Based Agent?

• A Utility-Based Agent makes choices on the basis of maximizing its overall satisfaction or
gain, assuming it evaluates the utility of different alternative outcomes.

Utility Function that Ranks Actions:

• It employs a utility function that scores possible outcomes in such a way that the agent
can rank or choose actions with highest utility.

2. Application and Use Cases:

When Utility-Based Agents Are Very Helpful?

• Robotics: It should find pathways with minimum energy consumption


• Finance: Algorithms of trade decide which strategies yield most profit when the risk is
controlled
• Game AI: It should select actions with maximum winning chances, and it should select
the best strategy

5. Learning Agents
Learning Agents improve performance with experience and act based on feedback from the
environment.

How Learning Agents Improve Over Time:

They use feedback to tune up strategies through:

• Reinforcing learning
• Supervised learning in decision-making

Types of Learning Agents

Learning Agents consist of:

• Learning Element: Behavior modification based on feedback


• Performance Element: Actions depending on present knowledge
• Critic: Provides performance feedback
• Problem Generator: Suggests new learning environments

These components enhance the agent's performance and flexibility.

Examples and Benefits of Learning Agents

Example:

• Recommendation Systems: Netflix and Amazon use the concept to recommend items to their
customers according to user's activity

Benefits:

• Flexibility: The model adapts to changing situations


• Better Decision Making: The optimization of action through time
• Autonomy: Self-running models that learn continuously

Learning Agents are necessary in complex scenarios because they can learn and adapt.

THE END

You might also like