Artificial Intelligence: A Modern Approach (4th Edition)#
Chapter 2: Intelligent Agents#
What is an Intelligent Agent?#
- An intelligent agent refers to a program or robot that can autonomously perform tasks.
 
Components of an Intelligent Agent#
- Perception: An intelligent agent perceives the state of the environment through sensors.
- Visual sensors: such as cameras.
 - Audio sensors: such as microphones.
 - Tactile sensors: such as touchscreens and force sensors.
 - Geolocation sensors: such as GPS.
 
 - Reasoning: An intelligent agent makes inferences and judgments based on the perceived information.
- Logical reasoning: deriving conclusions using logical rules.
 - Probabilistic reasoning: making inferences using probability statistics.
 - Machine learning: making inferences by learning and acquiring knowledge.
 
 - Action: An intelligent agent changes the environment by executing operations.
- Actuators: such as motors and robotic arms.
 - Communication devices: such as Wi-Fi and Bluetooth.
 
 
Classification of Intelligent Agents#
- Simple reflex agents: They directly execute actions based on current perceptions.
 - Model-based agents: They build an internal model from observations of the environment to better perform actions.
- Model: an abstract description of the environment.
 - Uses: predicting environmental changes, planning actions.
- Environmental model: maps, scenes, etc.
 - Action model: how to execute tasks.
 
 
 - Learning agents: They improve performance through learning, including model-based learning and model-free learning.
- Model-based learning: learning using an environmental model.
- Supervised learning: learning from labeled data.
 - Reinforcement learning: learning through rewards and punishments.
 
 - Model-free learning: learning without using an environmental model, directly from interactions.
- Unsupervised learning: learning by discovering patterns in data.
 - Deep learning: learning through simulating neural networks.
 
 
 - Model-based learning: learning using an environmental model.
 - Autonomous agents: They can set goals, make plans, and self-evaluate and adjust.
- Goals: autonomous, long-term objectives.
- Long-term goals: achieving specific states or completing specific tasks.
 - Short-term goals: specific plans to achieve long-term goals.
 
 - Plans: sequential actions taken to achieve goals.
- Plan formulation: creating feasible plans.
 - Plan execution: executing plans and continuously adjusting.
 
 
 - Goals: autonomous, long-term objectives.