INTRODUCTION TO

ARTIFICIAL
INTELLIGENCE
PGS. TS. Nguyễn Thanh Bình
ENVIRONMENT TYPES
Fully observable (vs. partially observable)
Deterministic (vs. stochastic)
Episodic (vs. sequential)
Static (vs. dynamic)
Discrete (vs. continuous)
Single agent (vs. multi-agent)
ENVIRONMENT TYPES
Fully observable (vs. partially observable): an agent’s sensors give it access to
the complete state of the environment at each point in time.
ENVIRONMENT TYPES
Deterministic (vs. stochastic): the next state of the environment is completely
determined by the current state and the action executed by the agent. (If the
environment is deterministic except for the actions of other agents, the
environment is strategic). The stochastic environment is random in nature which
is not unique and cannot be completely determined by the agent.
Examples:
Chess : there would be only a few possible moves for a coin at the current
state and these moves can be determined.
Self-Driving Cars: the actions of a self-driving car are not unique, it varies time
to time.
ENVIRONMENT TYPES
Episodic (vs. sequential):
In an episodic task environment, the agent’s experience is divided into atomic
episodes. Each episode consists of the agent perceiving and then performing
a single action. The choice of action in each episode depends only the episode
itself.
In sequential environment, the current decision could affect all future
decisions, such as chess and taxi driving.
ENVIRONMENT TYPES
Static (vs. dynamic): the environment is unchanged while the agent is
deliberating.
Taxi driving is dynamic: the other cars and taxi itself keep moving while the
driving algorithm decides what to do next.
Crossword puzzles are static.
The environment is semi-dynamic if the environment itself does not change with
the passage of time by the agent’s performance score does.
For example: chess when played with a clock is semi-dynamic.
ENVIRONMENT TYPES
Discrete (vs. continuous): A limited number of distinct, clearly defined percepts
and actions.
Example:
Chess has a discrete set of percepts and actions.
Taxing driving is a continuous state and continuous-time problem.
ENVIRONMENT TYPES
Single-agent (vs. multi-agent): an agent operating by itself in an environment.
Example: an agent solving a crossword puzzle by itself is clearly in a single-agent
environment, whereas an agent playing chess is a two-agent environment.
AGENT FUNCTIONS
AND PROGRAMS
An agent is completely specified by the agent function mapping percept
sequences to actions.
The job of AI is to design the agent program that implements the agent
function mapping percept to actions.
One agent function is rational.
Aim: find a way to implement the rational agent function concisely.
AGENT FUNCTIONS
Take the current percept as input from the sensors (since nothing moreis
available from the environment) and return an action to the actuators.
Different from the agent function that takes the entire percept history.
If the agent’s actions depend on the entire percept sequence, the agenthas to
remember the percepts.
TABLE-DRIVEN AGENT
To keep track of the percept sequence and then use it to index into a table of
actions to decide what to do next.
To build a rational agent in this way, one must construct a table that contains
the appropriate action for every possible percept sequence
TABLE-DRIVEN AGENT
Cons:
Huge table
Take a long time to build the table
No autonomy
Need a long time to learn the table entries
VACUUM-CLEANER AGENT

Four basic kinds of agent programs:

Simple reflex agents
Model-based reflex agents
Goal-based agents
Utility-based agents
SIMPLE REFLEX AGENT
SIMPLE REFLEX AGENT
MODEL-BASED REFLEX AGENTS
MODEL-BASED REFLEX AGENTS
GOAL-BASED AGENT
UTILITY-BASED AGENTS
LEARNING AGENTS
SOLVING PROBLEMS
BY SEARCHING
Problem-solving agents
Problem types
Problem formulation
Examples
Basic search algorithms
PROBLEM-SOLVING AGENTS
PROBLEM-SOLVING AGENTS
On holiday in Vietnam; at the moment in Nha Trang.
Flight leaves tomorrow in Hochiminh city.
Formulate goal: be in Hochiminh city.
Formulate problem:
States: Various provinces.
Actions: drive between provinces.
Find solutions: sequence of provinces.
PROBLEM TYPES
Deterministic, fully observable: single-state problem. Agents know exactly
which state it will be in; solution is a sequence.
No-observable: sensorless problem (conformant problem). Agent may have no
idea where it is and solution is a sequence.
PROBLEM TYPES
Nondeterministic and/or partially observable: contingency problem: percepts
provide new information about the current state often interleave search,
execution
Unknown state space: exploration problem
EXAMPLE: VACUUM WORLD

Single-state, start in state 5.

Solutions?
EXAMPLE: VACUUM WORLD

Single-state, start in state 5.

Solutions?
[Right,Suck]
EXAMPLE: VACUUM WORLD

Sensorless:
Start in {1, 2, .., 6, 7, 8}
Right goes to {2, 4, 6, 8}
Solutions?
EXAMPLE: VACUUM WORLD
Contingency problems:
Nondeterministic: Suck may
dirty a clean carpet.
Partially observable:
location, dirty at current
location.
Percept: [L,Clean], i.e., start
in the state 5 or 7.
Solutions?
EXAMPLE: VACUUM WORLD
Contingency problems:
Nondeterministic: Suck may
dirty a clean carpet.
Partially observable:
location, dirty at current
location.
Percept: [L,Clean], i.e., start
in the state 5 or 7.
Solutions?
[Right, if Dirty then Suck]
SINGLE-STATE PROBLEM
FORMULATION
One problem can be determined by the following items:
Initial state: where the agent starts in, e.g., “at Nha Trang”
Actions or successor function S(x) = set of action-state pairs.
Goal test: explicit/implicit to determine whether a given state is a goal state.
Path cost: to show a numeric cost to each path, e.g., sum of distances,
number of actions executed,...
C(x,a,y) = step cost (non-negative)
SINGLE-STATE PROBLEM
FORMULATION
A solution of the problem is a sequence of necessary actions from the initial
state to achieve the goal state.
Solution quality is measured by the path cost function and an optimal solution
has the lowest path cost among all possible solutions.
SELECTING A STATE SPACE
Real world is complicated. State space must be abstracted for problem solving.
Abstract state = a set of real states.
Abstract action = a complex combination of real actions.
Abstract solution = a set of real paths that are solutions in the real world.
VACUUM WORLD
STATE SPACE GRAPH
States?
Actions?
Goal test?
Path cost?
VACUUM WORLD
STATE SPACE GRAPH
States?
integer dirt and robot location
Actions?
Left, Right, Suck
Goal test?
No dirt at all locations
Path cost?
1 per action
THE 8-PUZZLE

States?
Actions?
Goal test?
Path cost?
THE 8-PUZZLE
States? locations of tiles
Actions? move blank left, right, up,
down
Goal test? a given goal state
Path cost? 1 per move
Optimal solution of n-Puzzle family
is NP-hard!!!
TREE SEARCH ALGORITHMS
After formulating the problem, we have to solve it. It can be done by a search
through the state space.
Basic idea:
TREE SEARCH ALGORITHMS
TREE SEARCH ALGORITHMS

General tree-search algorithm

GENERAL TREE-SEARCH
ALGORITHMS
Implementation: States vs Nodes
A state is a representation of a physical
configuration
A node is a data structure constituting part of
a search tree, including: state, parent node,
action, path-cost and depth (the number of
steps along the path from the initial state).
GENERAL TREE-SEARCH
ALGORITHMS
The “Expand” functions creates new nodes, filling in the various fields and
using the “SuccessorFn” of the problem to create the corresponding states.
A fringe is a collection of nodes that have been generated but not
expanded.
TREE SEARCH ALGORITHMS
A search strategy is defined by picking the order of node expansion.
Strategies are evaluated along the following dimensions:
completeness: does it always find a solution if one exists?
time complexity: number of nodes generated
space complexity: How much memory is needed to perform the search
optimality: does it always find a least-cost solution?
Time and space complexity are measured in terms of
b: maximum branching factor of the search tree
d: depth of the least-cost solution
m: maximum depth of the state space