Artificial Intelligence

Dr. Hashem Tamimi

Palestine Polytechnic University
2024
Course Outline
• Introduction to AI

• Intelligent searching and game playing

• Genetic Algorithms

• Probabilistic models

• Handling Uncertainty

• Introduction to machine learning

• Selected Topics
 References

• G. F. Luger, Artificial Intelligence: Structures and

Strategies for Complex Problem Solving,
5th Ed., Addison Wesley, 2005.

• S. Russell & P. Norvig, Artificial Intelligence:

A Modern Approach, 2nd Ed., Prentice Hall, 2002.
Prerequisites

• Data Structures
• Computer Algorithms
• Logic
• Probability
What is AI

AI:
The branch of computer science concerned with the automation of the
intelligent (human/animal/living) behavior.

What is intelligence in AI?

What is intelligence?
• Turing Test
Alan Turing's 1950 paper Computing Machinery and
Intelligence

human

interrogator

computer
Agents
Agents
• Artificial intelligence is defined as the study of rational agents.

• An agent is anything that can be viewed as:

• perceiving its environment through sensors and
• acting upon that environment through actuators
Agents
• Examples of Agent:
• A Robotic agent has Cameras and infrared range finders which act as
sensors and various motors acting as actuators.
• A software agent has Keystrokes, file contents, received network
packages (which act as sensors) and displays on the screen, files, sent
network packets (acting as actuators).
• A Human-agent has eyes, ears, and other organs which act as
sensors, and hands, legs, mouth, and other body parts acting as
actuators.
Agents

• A rational agent is said to perform the right things.

• A Performance measure is used that defines the success criterion.
Agents

• Example: Self driving car

• Performance: Safety, time, comfort
• Environment: Roads, other vehicles, road signs, pedestrian
• Sensors: Camera, GPS, speedometer, accelerometer
• Actuators: Steering, accelerator, brake, signal
Agents
• Example 2:
• The game of sokoban
• Performance: ____________
• Environment: _____________
• Sensors: _________________
• Actuators: ________________
Types of agents
Types of agents
Simple reflex agent

Model-based agent

Goal-based agent

Utility-based agent

Learning agent
Types of agents
Simple reflex agent
• Simple Reflex Agents:
• Operates based on a set of condition-action rules, Model-based agent
also known as "if-then" rules or production rules.

• Does not consider the past percepts Goal-based agent

• Does not consider the consequences

Utility-based agent
of the action

• Does not consider any internal state Learning agent

representing the world.
Types of agents
Simple reflex agent
• Model-based agent
• A model-based agent continuously maintains an
Model-based agent
internal model or representation of the
environment. Goal-based agent
• The agent uses this model to make decisions by
considering different possible actions and their Utility-based agent
outcomes.
Learning agent
• It selects the action that leads to the best
outcome according to its goals or objectives.
Types of agents
Simple reflex agent
• Goal-based agent
• Build on top of Goal based agent. Model-based agent
• The objective (goal state) is clearly defined.
Goal-based agent
• There are different states (alternatives) that
can be reached by a set of actions
Utility-based agent
• the agent selects actions that it believes will
move it closer to achieving its objectives. Learning agent
Types of agents
Simple reflex agent
• Utility-based agent
• Consider the best actions to reach a set of Model-based agent
preferences (including reaching the goal)

• This is usually done using a utility function Goal-based agent

• Utility function represents the agent's

Utility-based agent
preferences or goals, capturing the
desirability of different outcomes. Learning agent
Types of agents
Simple reflex agent
• Learning-based agent
• This agent start without knowledge of how to
act to reach the goal or preference. Model-based agent
• The agent has a learning capability that allows it
to adjust its action.
Goal-based agent
• A performance measure is used to judge the
agent performance after each action
• Feedback from the performance measure Utility-based agent
will allow the agent to update its
actions to improve its performance.
Learning agent
State Representation and
search space
Sokoban game
• The elements of the game:
• Player
• Walls
• Boxes
• Storage locations (equal to the boxes)
• Condition:
• Boxes can only be pushed forwards
(they cannot be pulled)
• Player can push one box only.
• player moves up, down, left or right
• Player cannot penetrate boxes or walls
• Goal: Each box is in a storage location.
Sokoban game
• Representation
• 2D array
• Empty space: 0
• Wall : 1
• Box 2
• Storage location 3
• Box in a storage location 4
• Person 5

• Q: how to represent the goal state?

Sokoban game
• Representation
• 2D array
• Empty space: 0
• Wall : 1
• Box 2
• Storage location 3
• Box in a storage location 4
• Person 5

• Q: how to represent the goal state?

Sokoban game
• Representation
• 2D array
• Empty space: 0
(S1) (S2)
• Wall : 1
• Box 2
• Storage location 3
• Box in a storage location 4
• Person 5

• Q: which of the four states are valid? (S3) (S4)

Sokoban search space
Sokoban search space

Q: How many states are there ?

Search Space
• Search space

• Start State

• Goal State

• Action

• Solution
The branching factor

• Tic-Tac-Toe: 4
• Chess: 35
• Go: 200
• Arimaa: 17,000
Searching
Searching
• A search problem consists of:
• A state space

“N”
• A successor function
(with actions, costs)

“E”
• A start state and a goal test

• A solution is a sequence of actions (a plan) which transforms the start

state to a goal state
State Space Sizes?
• World state:
• Agent positions: 120
• Food count: 30
• Ghost positions: 12
• Agent facing: NSEW

• How many
• World states?
120x(230)x(122)x4
• States for pathing?
120
• States for eat-all-dots?
120x(230)
 State Space Graphs
• State space graph: A mathematical
representation of a search problem
• Nodes are (abstracted) world configurations
• Arcs represent successors (action results)
• The goal test is a set of goal nodes (maybe only one)

• In a state space graph, each state occurs only

once!

• We can rarely build this full graph in memory

(it’s too big), but it’s a useful idea
Depth-First Search
 Depth-First Search
Strategy: expand a a G
deepest node first b c

Implementation: e
d f
Fringe is a LIFO stack S h
p q r

d e p

b c e h r q

a a h r p q f

p q f q c G

q c G a

a
Search Algorithm Properties
• Infinite branch!
 Search Algorithm Properties
• Complete: Guaranteed to find a solution if one exists?
• Optimal: Guaranteed to find the least cost path?
• Time complexity?
1 node
• Space complexity? b
… b nodes
b2 nodes
• Cartoon of search tree:
m tiers
• b is the branching factor
• m is the maximum depth
• solutions at various depths
bm nodes
• Number of nodes in entire tree?
• 1 + b + b2 + …. bm = O(bm)
 Depth-First Search (DFS) Properties
• What nodes DFS expand?
• Some left prefix of the tree. 1 node
• Could process the whole tree! b
… b nodes
• If m is finite, takes time O(bm)
b2 nodes
• How much space does the fringe take? m tiers
• Only has siblings on path to root, so O(bm)

• Is it complete?
• m could be infinite, so only if we prevent bm nodes
cycles (more later)

• Is it optimal?
• No, it finds the “leftmost” solution,
regardless of depth or cost
Breadth-First Search
 Breadth-First Search
Strategy: expand a a G
shallowest node first b c
Implementation: Fringe e
d f
is a FIFO queue S h
p q r

d e p
Search
b c e h r q
Tiers
a a h r p q f

p q f q c G

q c G a

a
 Breadth-First Search (BFS) Properties
• What nodes does BFS expand?
• Processes all nodes above shallowest solution 1 node
b
• Let depth of shallowest solution be s … b nodes
• Search takes time O(bs) s tiers
b2 nodes

• How much space does the fringe take? bs nodes

• Has roughly the last tier, so O(bs)

• Is it complete? bm nodes
• s must be finite if a solution exists, so yes!

• Is it optimal?
• Only if costs are all 1 (more on costs later)
Quiz: DFS vs BFS
• In the previous example what are the visited nodes of the tree using
BFS?
• Using DFS?
• What is the solution path from root to goal ?
Another example
DFS vs BFS
• When will BFS outperform DFS?

• When will DFS outperform BFS?

Quiz

• Depth First Search

• A → E→ i→m → n →o
→p→I →H→g

• Breadth first search

• A→
e→f→b
→i→j→c→
m→n→g
Blind Search vs Intelligent Search
• Blind Search:
• Include BFS and DFS.
• During the search process, the agent has no clue if it is far or near a goal!

• Heuristic search:
• Eg: Best First Search
• Eg: A* search
• The agent has some hint (heuristic) how far it is from goal
• Heuristic search is intelligent search
Heuristic Search

The agent has some hint (heuristic) how far it is from goal
Heuristic Search
A heuristic function:
A function that estimates how close a state is to a goal
Designed for a particular search problem
Examples: Manhattan distance, Euclidean distance for pathing

10

5
11.2
Heuristic Search

• A heuristic function: • The value v reflects how far the

• Takes one state as input state s is from the goal state
• Provide a numeric value as
output

a state s h(s) a numeric

value v
Travel Salesman Problem

Traveling Salesman Problem:

Complexity of the exhaustive

search is: (n-1)!

Time consuming when n is large.

Travel Salesman Problem
Travel Salesman Problem
• Heuristic function : Select the nearest city next (Greedy)
Heuristic Search
• Consider different heuristics of
• 8-puzzle problem:
Heuristic Search
• Consider different heuristics of
8-puzzle problem:

h1: # of misplaced tiles ignores

the distances the tiles must be
moved.
misplacedTilesHeuristic
Heuristic Search
Heuristic Search
• Tic-Tac-Toe:
• Complexity of exhaustive search
= 9x8x7,... = 9!.
• With symmetry reduction:
• 12x7!
Heuristic Search
• Tic-Tac-Toe (continued):
Heuristic: Move to the board in
which X has the most winning lines.
Heuristic Search
Admissibility
• An admissible heuristic is a heuristic that is guaranteed to find the
shortest path from the current state to the goal state.
• To check if a heuristic function h is admissible or not, we need to
check that:
ℎ 𝑠 ≤ ℎ∗ 𝑠 ∀𝑠 ∈ 𝑆, where 𝑆 is the state space
• ℎ∗ 𝑠 is an optimal heuristic function
Quiz
• Which heuristic is admissible?
Consistent
• Consistent or monotonic
• h is consistent if: n
1. h(goal)=0
2. h(s) ≤ c(s,n) + h(n)
For all s belong to S S
Where n is any direct neighbors state s g

Note
If H:consistent=>h(admissible)
Consistent
• Consistent or monotonic
• h is consistent if: n
1. h(goal)=0
2. h(s) ≤ c(s,n)+h(n)
For all s belong to S S
Where n is any direct neighbors state s
g

Note
If H:consistent ➔ h:admissible
But the opposite is not true
 A* search
• h(s) is the heuristic value for state s
• g(s) is the distance from the start to the state
• f(s)=h(s)+g(s)
• h(s): admissible for all s
Quiz
Solve using:

• Breadth First Search

• Depth First Search

• Best First Search

• Best First Search including the cost on edges

•
Quiz
• Which heuristic is monotonic?
MiniMax
• Nim-game
MiniMax
pruning