INTRO TO ARTIFICIAL INTELLIGENCE

SEARCH
AHMED EZZAT LABIB
HELWAN UNIVERSITY
 HILL CLIMBING SEARCH

• QUEUEINGFN IS SORT-BY-H
• ONLY KEEP LOWEST-H STATE ON OPEN LIST

• BEST-FIRST SEARCH IS TENTATIVE (NOT CERTAIN OR FIXED)

• HILL CLIMBING IS IRREVOCABLE (NOT ABLE TO BE CHANGED; FINAL)
• FEATURES
• MUCH FASTER
• LESS MEMORY
• DEPENDENT UPON H(N)
• IF BAD H(N), MAY PRUNE AWAY ALL GOALS
• NOT COMPLETE
EXAMPLE
EXAMPLE
 HILL CLIMBING – THE 8-PUZZLE PROBLEM
A 3 1 2 1 2
h=5 4 5 8 3 4 5
6 7 6 7 8
Goal State
h is the number of incorrect
positions (or the misplaced
tiles) in the puzzle

Path = { }
Open = {(A, 5)}
Current = {A}
 HILL CLIMBING – EXAMPLE 1 2
A 3 1 2 3 4 5
h=5 4 5 8 6 7 8
6 7 Goal State
B C D
3 1 2 3 1 2 3 1 2
h=6 4 5 8 h=5 4 8 h=4 4 5 8
6 7 6 5 7 6 7

Path = {A}
Open = {(D, 4), (B, 6), (C, 5)}
Current = {D}
 HILL CLIMBING – EXAMPLE 1 2
A 3 1 2 3 4 5
h=5 4 5 8 6 7 8
6 7 Goal State
B C D
3 1 2 3 1 2 3 1 2
h=6 4 5 8 h=5 4 8 h=4 4 5 8
6 7 6 5 7 6 7
E F
3 1 2 3 1 2
Path = {A, D} h=3 4 5 h=5 4 5 8
Open = {(E, 3), (F, 5)} 6 7 8 6 7
Current = {E}
 HILL CLIMBING – EXAMPLE 1 2
A 3 1 2 3 4 5
h=5 4 5 8 6 7 8
6 7 Goal State
B C D
3 1 2 3 1 2 3 1 2
h=6 4 5 8 h=5 4 8 h=4 4 5 8
6 7 6 5 7 6 7
E F
3 1 2 3 1 2
Path = {A, D, E} h=3 4 5 h=5 4 5 8
Open = {(H, 2), (G, 4) , (I, 4)} 6 7 8 6 7
Current = {H}
G H I
3 1 3 1 2 3 1 2
h=4 4 5 2 h=2 4 5 h=4 4 5 8
6 7 8 6 7 8 6 7
 1 2
HILL CLIMBING – EXAMPLE 3 4 5
G H I 6 7 8
3 1 3 1 2 3 1 2 Goal State
h=4 4 5 2 h=2 4 5 h=4 4 5 8
6 7 8 6 7 8 6 7

h=3 J h=1 K h=3 L M h=3

3 1 2 3 1 2 3 2 3 1 2
4 5 4 5 4 1 5 4 7 5
6 7 8 6 7 8 6 7 8 6 8

Path = {A, D, E, H}
Open = {(K, 1), (J, 3) , (L, 3) , (M, 3)}
Current = {K}
 HILL CLIMBING – EXAMPLE
G H I
3 1 3 1 2 3 1 2
h=4 4 5 2 h=2 4 5 h=4 4 5 8
6 7 8 6 7 8 6 7

h=3 J h=1 K h=3 L M h=3

3 1 2 3 1 2 3 2 3 1 2
4 5 4 5 4 1 5 4 7 5
6 7 8 6 7 8 6 7 8 6 8

h=2 N h=0 O P h=2

3 1 2 1 2 3 1 2 Path = {A, D, E, H, K}
6 4 5 3 4 5 4 5 Open = {(O, 0), (N, 2) , (P, 2)}
7 8 6 7 8 6 7 8 Current = {O}
Goal State Goal State
 HILL CLIMBING SEARCH

• YOU CAN READ HILL CLIMBING SEARCH FROM ANY

TEXT BOOK
• YOU CAN SEARCH ABOUT HILL CLIMBING SEARCH
WITH ANOTHER EXAMPLE (THE 4-QUEEN OR 8-QUEEN
PROBLEM)
• YOU CAN SEARCH FOR ANY VIDEO THAT TALK ABOUT
HILL CLIMBING SEARCH WITH THE 8-PUZZLE PROBLEM
FOR MORE DESCRIPTION.
COMPARISON OF SEARCH TECHNIQUES

DFS BFS UCS IDS Best HC

Complete N Y Y Y N N
Optimal N N Y N N N
Heuristic N N N N Y Y
Time bm bd+1 bm bd bm mn
Space bm bd+1 bm bd bm b
 INFORMED SEARCH – A*

• IDEA: AVOID EXPANDING PATHS THAT ARE ALREADY EXPENSIVE

• EVALUATION FUNCTION F(N) = G(N) + H(N)
• G(N) = COST SO FAR TO REACH N
• H(N) = ESTIMATED COST FROM N TO GOAL
• F(N) = ESTIMATED TOTAL COST OF PATH THROUGH N TO GOAL
• BEST FIRST SEARCH HAS F(N)=H(N)
• UNIFORM COST SEARCH HAS F(N)=G(N)
 INFORMED SEARCH – A*

• QUEUEINGFN IS SORT-BY-F ( F(N) = G(N) + H(N) )

• NOTE THAT UCS AND BEST-FIRST BOTH IMPROVE SEARCH

• UCS KEEPS SOLUTION COST LOW
• BEST-FIRST HELPS FIND SOLUTION QUICKLY

• A* COMBINES THESE APPROACHES

A* – EXAMPLE
A* – EXAMPLE
A* – EXAMPLE
A* – EXAMPLE
A* – EXAMPLE
A* – EXAMPLE
A* – EXAMPLE
A* – EXAMPLE
 A* SEARCH – EXAMPLE 2
g(n)=0

23
 A* SEARCH – EXAMPLE 2

g(n)=0+140 g(n)=0+118 g(n)=0+75

 A* SEARCH – EXAMPLE 2
g(n)=140

g(n)=140+140
g(n)=140+80

g(n)=140+99 g(n)=140+151

25
 A* SEARCH – EXAMPLE 2

g(n)=0+140
g(n)=220

g(n)=140+99

g(n)=220+146 g(n)=220+97 g(n)=220+80

 A* SEARCH – EXAMPLE 2

g(n)=239
g(n)=220

g(n)=220+80

g(n)=239+211 g(n)=220+146 g(n)=220+97

27
 A* SEARCH – EXAMPLE 2

g(n)=317

g(n)=317+101
g(n)=317+138

28
 A* SEARCH

• YOU CAN READ A* SEARCH FROM $1< TEXT BOOK

• YOU CAN SEARCH FOR ANY VIDEO THAT TALK ABOUT

A* SEARCH FOR MORE DESCRIPTION.
 COMPARISON OF SEARCH TECHNIQUES

DFS BFS UCS IDS Best HC Beam A*

Complete N Y Y Y N N N Y

Optimal N N Y N N N N Y

Heuristic N N N N Y Y Y Y

Time bm bd+1 bm bd bm bm nm bm
Space bm bd+1 bm bd bm b bn bm
 PROBLEM – HILL CLIMBING SEARCH
• CONSIDER THE INITIAL STATE OF THE 4-QUEENS PROBLEM SHOWN BELOW.
• Consider the hill-climbing with number of conflicts heuristic
to find a valid state of the 4-queens problem.
• The heuristic value of a state is the number of distinct
queen pairs that can attack each other.
• The successors of a state are all possible states generated
by moving a single queen to another square in the same
column.
• Hill-climbing would at each step take the successor with
the smallest number of conflicts.
• DRAW THE TREE OF STATES THAT HILL-CLIMBING USING THE MINIMUM CONFLICT HEURISTIC EXPLORES
FROM THE INITIAL STATE TO ALL POSSIBLE EXPLORATION PATHS OF HILL-CLIMBING.
• ALSO WRITE THE OPEN LIST FOR THIS PROBLEM.

• DEADLINE: 11/11/2023 - THE REPORT MUST BE IN A HAND WRITTEN FORMAT