UNIVERSITY OF OXFORD | MSC
ARTIFICIAL INTELLIGENCE
CS6101: FOUNDATIONS OF AI & MACHINE LEARNING (TERM 1)
COMPREHENSIVE LECTURE SERIES & TEACHING NOTES (2026 EDITION)
MODULE 1: THE RATIONAL AGENT FRAMEWORK
The Oxford curriculum shifts the focus from "simulating humans" to the
mathematical formalization of rationality.
1.1 Defining the Rational Agent
A Rational Agent is a system that perceives its environment and takes actions that
maximize its expected performance measure.
• The PEAS Analysis (Task Environment Specification):
o P (Performance Measure): The objective criteria for success (e.g.,
minimizing fuel consumption in a self-driving car).
o E (Environment): The world the agent operates in (e.g., UK
motorways, pedestrians).
o A (Actuators): How the agent acts (e.g., steering, braking).
o S (Sensors): How the agent perceives (e.g., LIDAR, cameras, GPS).
1.2 Environment Properties
Understanding the environment determines the complexity of the AI algorithm:
• Fully Observable vs. Partially Observable: Does the agent see everything
relevant?
• Deterministic vs. Stochastic: Does an action always lead to the same
result?
• Static vs. Dynamic: Does the world change while the agent is "thinking"?
ARTIFICIAL INTELLIGENCE
CS6101: FOUNDATIONS OF AI & MACHINE LEARNING (TERM 1)
COMPREHENSIVE LECTURE SERIES & TEACHING NOTES (2026 EDITION)
MODULE 1: THE RATIONAL AGENT FRAMEWORK
The Oxford curriculum shifts the focus from "simulating humans" to the
mathematical formalization of rationality.
1.1 Defining the Rational Agent
A Rational Agent is a system that perceives its environment and takes actions that
maximize its expected performance measure.
• The PEAS Analysis (Task Environment Specification):
o P (Performance Measure): The objective criteria for success (e.g.,
minimizing fuel consumption in a self-driving car).
o E (Environment): The world the agent operates in (e.g., UK
motorways, pedestrians).
o A (Actuators): How the agent acts (e.g., steering, braking).
o S (Sensors): How the agent perceives (e.g., LIDAR, cameras, GPS).
1.2 Environment Properties
Understanding the environment determines the complexity of the AI algorithm:
• Fully Observable vs. Partially Observable: Does the agent see everything
relevant?
• Deterministic vs. Stochastic: Does an action always lead to the same
result?
• Static vs. Dynamic: Does the world change while the agent is "thinking"?
