Types of AI Agents
Understanding Different Kinds of Agents in Artificial
Intelligence
Er. Rudra Nepal
November 13, 2024
What is an AI Agent?
▶ An AI agent perceives its environment and takes actions to
achieve its goals.
▶ The agent senses its environment through percepts (inputs),
makes decisions, and performs actions.
▶ The way the agent behaves depends on how it processes its
environment and the rules or learning algorithms it follows.
Types of AI Agents
▶ Simple Reflex Agent
▶ Model-Based Reflex Agent
▶ Goal-Based Agent
▶ Utility-Based Agent
▶ Learning Agent
Simple Reflex Agent
▶ A simple reflex agent reacts to the current situation using
predefined rules.
▶ It doesn’t think ahead or remember past actions. It only
reacts to the current percept.
▶ **How it works**: Based on a condition in the environment,
it immediately performs a corresponding action.
▶ **Example**: A robot vacuum that moves forward until it
hits an obstacle, then turns and moves in another direction.
▶ **Rule**: If an obstacle is detected, turn left.
Model-Based Reflex Agent
▶ A model-based reflex agent remembers past states of the
world and uses this memory to make better decisions.
▶ Unlike the simple reflex agent, it doesn’t just react to the
current situation; it considers how the world has changed.
▶ **How it works**: It keeps track of what has happened before
and uses this information to update its model of the world.
▶ **Example**: A thermostat. It doesn’t just react to the
current temperature but remembers the last temperature and
adjusts the heating or cooling accordingly.
▶ **Rule**: If the temperature is too low (and it was also low
earlier), turn on the heater.
Goal-Based Agent
▶ A goal-based agent has specific goals it aims to achieve, and
it plans actions to reach those goals.
▶ Unlike reflex agents, it isn’t just reacting. It actively thinks
about how to achieve a desired outcome.
▶ **How it works**: It evaluates its options and chooses the
best sequence of actions to achieve its goal.
▶ **Example**: A GPS navigation system. The system doesn’t
just react to roads or traffic; it figures out the best route to a
destination.
▶ **Goal**: Find the fastest route from point A to point B.
Utility-Based Agent
▶ A utility-based agent chooses actions based on a measure of
”utility” or satisfaction.
▶ It doesn’t just aim to achieve a goal; it tries to maximize its
”happiness” or ”satisfaction” from different options.
▶ **How it works**: It evaluates possible actions according to
how much utility they provide and selects the one that gives
the highest utility.
▶ **Example**: A shopping assistant. It helps you choose the
best product by evaluating quality, price, reviews, etc., to find
the option that maximizes your satisfaction.
▶ **Utility**: Choose the product that gives the most value
(quality + price) based on your preferences.
Learning Agent
▶ A learning agent improves over time by learning from its
experiences.
▶ It starts with some basic knowledge or rules and then uses
feedback from the environment to improve its actions.
▶ **How it works**: It uses past experiences (correct or
incorrect actions) to update its knowledge or strategy,
improving its future performance.
▶ **Example**: A game-playing AI that gets better at playing
chess the more it plays, learning new strategies from its wins
and losses.
▶ **Learning Process**: Observe → Learn → Improve →
Perform better.
Summary
▶ Simple Reflex Agent: Reacts to current situations based on
predefined rules, with no memory or planning.
▶ Model-Based Reflex Agent: Remembers past states of the
world to make better decisions in the present.
▶ Goal-Based Agent: Chooses actions that bring it closer to a
specific goal.
▶ Utility-Based Agent: Chooses actions that maximize
satisfaction or ”utility.”
▶ Learning Agent: Learns from past experiences to improve
over time.
Thank You!
Any Questions?
