Artificial Intelligence

Hai Thi Tuyet Nguyen

1
Nội dung môn học
CHAPTER 1: INTRODUCTION (CHAPTER 1)
CHAPTER 2: INTELLIGENT AGENTS (CHAPTER 2)
CHAPTER 3: SOLVING PROBLEMS BY SEARCHING (CHAPTER 3)
CHAPTER 4: INFORMED SEARCH (CHAPTER 3)
CHAPTER 5: LOGICAL AGENT (CHAPTER 7)
CHAPTER 6: FIRST-ORDER LOGIC (CHAPTER 8, 9)
CHAPTER 7: QUANTIFYING UNCERTAINTY(CHAPTER 13)
CHAPTER 8: PROBABILISTIC REASONING (CHAPTER 14)
CHAPTER 9: LEARNING FROM EXAMPLES (CHAPTER 18)
2
CHAPTER 2: 2.1 Agents And Environments
2.2 The Concept Of Rationality
INTELLIGENT 2.3 The Nature Of Environments
2.4 The Structure Of Agents

AGENTS
2.5 Summary

3
2.1 Agents And Environments
● An agent is something that
○ perceives its environment through sensors
○ acts upon that environment through actuators.
● A human agent has eyes, ears, … for sensors and hands, legs, … for actuators.

4
2.1 Agents And Environments
● Agent’s percept: the agent’s inputs at any given instant.
● Agent’s percept sequence: the complete history of everything that the agent has
perceived.
● Agent’s behavior (~agent function): maps any percept sequence to an action.
● Agent program: the concrete implementation of the agent action

5
2.1 Agents And Environments
● Percepts: location and contents, e.g., [A, Dirty]
● Actions: Left, Right, Suck, NoOp

6
2.1 Agents And Environments

7
2.2 The Concept Of Rationality
● A rational agent chooses the action that maximizes the expected value of the
performance measure given the percept sequence.

8
2.3 The Nature Of Environments
2.3.1 Specifying the task environment
● To design a rational agent, we must specify the environment task.
● Task environment or PEAS (Performance, Environment, Actuators, Sensors):

9
2.3 The Nature Of Environments
2.3.2 Properties of task environments
● Fully observable, partially observable, unobservable
○ fully observable if the sensors detect all relevant aspects.
○ partially observable because of noisy and inaccurate sensors or parts of the state are missing from the
sensor data
○ unobservable if the agent has no sensors at all
● Single agent vs. multiagent
○ single-agent: an agent solving a crossword puzzle by itself
○ two- agent:
■ Competitive: chess
■ Cooperative: taxi-driving

10
2.3 The Nature Of Environments
2.3.2 Properties of task environments
● Deterministic vs. stochastic
○ Deterministic: if the next state of the environment is completely determined by the current state and the
action executed by the agent
○ Stochastic: otherwise; generally implies that uncertainty about outcomes is quantified in terms of
probabilities
● Episodic vs. sequential
○ Episodic: the agent’s experience is divided into atomic episodes; the next episode does not depend on
the actions taken in previous episodes.
○ Sequential: the current decision could affect all future decisions; e.g., chess and taxi driving

11
2.3 The Nature Of Environments
2.3.2 Properties of task environments
● Static vs. dynamic:
○ dynamic: if the environment can change while an agent is deliberating; e.g., taxi driving
○ static: otherwise; e.g., crossword puzzles
○ semi-dynamic: if the environment does not change with the passage of time but the agent’s performance
score does, e.g., chess, when played with a clock
● Discrete vs. continuous: applies to the state of the environment, to the way time is
handled, and to the percepts and actions of the agent
○ the chess environment has a finite number of distinct states, a discrete set of percepts and actions
○ taxi driving is a continuous-state and continuous-time problem, its actions are also continuous
● Known vs. unknown:
○ In a known environment, the outcomes for all actions are given.

12
2.4 The Structure Of Agents
● The job of AI is to design an agent program that implements the agent function
● This program runs on computing device with physical sensors, actuators: architecture

AGENT = ARCHITECTURE + PROGRAM

13
2.4 The Structure Of Agents
2.4.1 Agent programs
● They take the current percept as input and return an action.
● Four basic types in order of increasing generality
○ simple reflex agents
○ reflex agents with state
○ goal-based agents
○ utility-based agents

14
2.4 The Structure Of Agents
2.4.2 Simple reflex agents:
● the simplest type
● the agents select actions on the basis of the current percept

15
2.4 The Structure Of Agents
2.4.3 Model-based reflex agents
● The agents maintain internal state to track aspects of the world that are not evident in
the current percept

16
2.4 The Structure Of Agents
2.4.4 Goal-based agents
● the agent needs some sort of goal information
● the agent program combines the goal information with the model to choose right actions

17
2.4 The Structure Of Agents
2.4.5 Utility-based agents
● Goals provide a binary distinction between “happy” and “unhappy” states.
● A performance measure should score exactly how happy they would make the agent.
● An agent’s utility function: an internalization of the performance measure.

18
 2.4 The Structure Of Agents
2.4.6 Learning agents
● All agents can improve their performance through learning with 4 conceptual components:
○ learning element uses feedback and determines how the performance element should be modified to do
better => making improvements
○ performance element takes in percepts and decides on actions => selecting external actions
○ problem generator suggests actions that will lead to new and informative experiences

19