Uninformed Search

Punjab University College of Information Technology, Lahore

© AI
 Blind Search Algorithms
 Backtrack
 Depth First
 Breadth First
 Depth First with iterative
deepening

© AI
 Backtracking
 Backtracking is a technique for
systematically trying all paths through
the state space.
 Backtracking search begins at the start
state and pursues a path until is
reaches either a goal or a dead end.
 If it finds goal, it quits and returns the
solution
 If it reaches the dead end, it backtracks
to the most recent node on the path
having unexamined siblings.

© AI
 Backtracking
Three Lists that keep track of the algorithm:
 SL
 for state list (in the current path being tried). if
goal is found, SL contains the ordered list on
the solution path.
 NSL
 For new state list (contains nodes awaiting
evaluation).
 DE
 For dead ends (lists states whose descendants
have failed to contain a goal node).
 CS
 For current state (contains node currently
under consideration)
© AI
 Backtrack Algorithm
begin
SL := [Start], NSL:= [Start], DE:= [], CS:= Start
while NSL≠ [] do
begin
if CS = goal
then return SL
if CS has no child nodes (excluding those already on DE, SL, and NSL)
then begin
while SL is not empty and CS = first element of SL do
begin
add CS to DE
remove first element from SL
remove first element from NSL
CS:= first element of NSL
end
add CS to SL
end
else begin
place children of CS (except nodes already on DE, SL, or NSL) on NSL
CS:= first element of NSL
add CS to SL
end
end
return FAIL
end
© AI
Backtracking Example
A

B C D

E
F G

J
I
H

© AI
 Backtracking Example
Iteration# CS SL NSL DE
0 A [A] [A] []

1 B [BA] [BCDA] []
2 E [EBA] [EFBCDA] []
3 H [HEBA] [HIEFBCDA] []
4 I [IEBA] [IEFBCDA] [H]
5 F [FBA] [FBCDA] [EIH]
6 J [ JFBA] [JFBCDA] [EIH]
7 C [CA] [CDA] [BFJEIH ]
8 G [GCA] [GCDA] [BFJEIH ]

© AI
Breadth First Search Algorithm
 Breadth first search explores the space
in a level by level fashion.
 Only when there are no more states to
be explored at a given level does the
algorithm move on the next level.
 “Open” and “Closed” lists are used to
keep track of the search. Open is Like
NSL having list of nodes whose children
have not been examined. Closed
records states that have already been
examined. Closed is the union of DE
and SL.
© AI
 BFS Example
# Open Closed
1 [A] []
2 [BCD] [A]
3 [CDEF] [BA]
4 [DEFG] [CBA]
5 [EFG] [DBA]
6 …. …

© AI
 Breadth First Search
Analysis
 Suppose Branching Factor is “b”
 Search tree generates “b” nodes at
Level-1 and b2 at Level-2 and . . . bd at
Level d.
 Here we can see the exponential
complexity bound like O(bd)
 The memory requirements are a bigger
problem for breadth first search.

© AI
 Breadth First
Depth Nodes Time Memory

0 1 1 ms 100 bytes
2 111 1 second 11 Kb
4 11111 11 seconds 1 MB
6 106 18 min 111MB
8 108 31 hrs 11GB
10 1010 128 days 1 TB
© AI
 Depth First Search
 Depth first is a simplification of back
track algorithm
 The descendent states are added and
removed from the left end of OPEN.
 OPEN is a LIFO structure. The
organization of open as a stack directs
search towards the most recently
generated states

© AI
Depth First Example
# Open Closed
1 [A] []
2 [BCD] [A]
3 [EFCD] [BA]
4 [HIFCD] [EBA]
5 [IFCD] [HEBA]
6 [FCD] [IHEBA]
7 [JCD] [FIHEBA]
8 [CD] [JFIHEBA]
9 …. …….
© AI
 Depth-First with iterative
Deepening
 The depth bound forces a failure on a
search path.
 This causes a breadth–like sweep of the
search at that depth level.
 This algorithm increases the bound level
in each iteration.
 Because the algorithm searches the
space in level–by–level fashion it is
guaranteed to find a shortest path to a
goal.

© AI
 Conclusion
 Breadth First
– Because it always examines all the nodes in
level by level fashion, it always finds the
shortest path to the goal
– In bad branching factor (states with high
average number of descendants) the
combinatorial explosion may prevent the
algorithm from finding a solution using the
available space.
– The space utilization is measured in terms
of the number of open states on OPEN is an
exponential function of the length of the
path at any time.
© AI
 Conclusion
 Depth First
– Gets quickly into a deep search space. If it
is known that the solution will be long,
depth first search will not waste time
finding it.
– It can get lost deep in the graph missing
shorter paths to the goal
– It might even get stuck in an infinitely long
path that does not lead to a goal.

© AI
 Heuristics
 Heuristics may be defined in a variety
of ways:
 The study of the methods and rules of
discovery and invention (Polya,1945)
 Heuristics may be viewed as:
 Rules for choosing those branches in a state
space that are most likely to lead to an
acceptable problem solution
 Heuristics are loosely described as rules
of thumb.

© AI
 Heuristics
 Two basic situations for the use of
Heuristics:
 An inherently ambiguous problem, e.g.
medical diagnosis system
 The cost of finding a solution by exhaustive
methods like depth first search, is
prohibitive
 A heuristic is fallible: it is only an
informed guess of next step to be taken
in solving a problem
 A heuristic can lead a search algorithm
to a sub-optimal solution or fail to find
any solution © AI
 Tic-Tac-Toe Problem
 Consider the example of state space
 9 nodes at the first level, 8 nodes of each
node at level–2 =9x8x7 . . 2 (9!=326880)
nodes to search
 Need to reduce the number of nodes or
the size of the state space
 Analyze carefully the rules of playing
Tic-Tac-Toe or view as expert to setup
heuristics

© AI
State Space Reduced by Symmetry!

© AI
 Tic-Tac-Toe problem
solving using Heuristic
 A simple heuristic can eliminate search
entirely:
 We may move to the board in which “X” has
the most winning lines.
 In case of states with equal numbers of
potential wins, take the first such state
found.
 In case “X” takes the center of the grid,
we can eliminate other alternates along
with all of its descendants to 2/3 of the
space. © AI
Tic-Tac-Toe problem solving
using Heuristic
Heuristics for State Space for TIC-TAC-TOE

X X

3 4 2

Dashed lines shows the number of potential wins at level-1

© AI
Tic-Tac-Toe using Heuristic
 Best-First Search
A-5

D-6
B-4 C-4

E-5 F-5 G-4 H-3 I J

K L N Q R
M O-2 P-3

S T
© AI
 Best-First Search
• Depth first Algorithm uses LIFO style list to
implement search
• Breadth First Search uses FIFO style list to
search
• Best-first is an heuristic search that uses a
priority queue to order the states on the
OPEN

© AI
 Best-First Search
Evaluate Opened Closed
[A5] []
A5 [B4,C4,D6] [A5]
B4 [C4,E5,F5,D6] [B4,A5]
C4 [H3,G4,E5,F5,D6] [C4,B4,A5]
H3 [O2,P3,G4,E5,F5,D6] [H3,C4,B4,A5]
O2 [P3,G4,E5,F5,D6] [O2,H3,C4,B4,A5]
P3 The solution

© AI
 Best-First Search
A-5

D-6
B-4 C-4

E-5 F-5 G-4 H-3 I J

K L N Q R
M O-2 P-3

S T
© AI
 Heuristic Measures for
8 Puzzle Problem
 The simple heuristic count the tiles out
place in each state when it is compared
with the goal.
 Sum all the distances by which the tiles
are out of place, one for each square a
tile must be moved to reach its position
in the goal state

© A.I
Heuristic Measures for
8 Puzzle Problem

2 8 3
1 6 4 h(n) = No. of Tiles out of place = 4
7 5

2 8 3 h(n) = Sum of distances out of place = 5

© A.I
A Possible Result of a Heuristic
Search Procedure

© A.I
Heuristic Evaluation Function

 If two states have same or nearly same

© A.I
 Heuristic Evaluation Function

2 8 3
g(n) = 0 1 6 4
7 5 h(n) = 4

2 8 3 2 8 3 2 8 3
1 6 4 1 6 4 1 6 4
g(n) = 1 7 7 5 7 6 5 7 5 5
h(n) = 5 h(n) = 3 h(n) = 5

Values of f(n): 6 4 6
© A.I
Heuristic Search using the
evaluation function

© A.I
 Characteristics of Heuristics
We can evaluate the behavior of heuristics on
the basis of following characteristics:
 Admissible:(gives shortest path)
 Informed:(has more accurate information
for next step to be taken)
 Monotone:(min-cost to find shortest path)

© A.I
 Admissibility
 A search algorithm is admissible if, for any
graph, it always terminates in the optimal
solution path.
 If the ‘h’ function never overestimates the
cost to reach the goal, then ‘h’ is called an
admissible heuristic
 Admissible functions are optimistic, as they
always think the cost to the goal is less
than it actually is!

© A.I
 Informedness
 For two heuristics ‘h1’ & ‘h2’, if
h1(n) < h2(n)
for all states ‘n’ in the search space, ‘h2’
is said to be more informed than ‘h1’
 A ‘more’ informed heuristic expands less
states than a relatively ‘less’ informed
heuristic

© A.I
 Monotonicity
 A heuristic function ‘h’ is monotonic if
 For all states ni and nj where nj is
descendent of ni
h(ni) - h(nj)  cost (ni,nj)
where cost (ni,nj) is the actual cost of
going from state ni to nj.
 The heuristic evaluation of the goal state is
zero or h(goal)=0.
 A monotonic heuristic consistently finds the
minimal path to each state it encounters in
the search.
© A.I
 Heuristic Search Algorithms
 Heuristic Search Algorithms
 Best-First Search
 Maintains a priority queue to switch to
heuristically promising paths
 A* Search
 Uses algorithm A (best-first search +
evaluation function f(n)) with a constraint
that h(n) never overestimates h*(n) -- the
actual cost to reach the goal from any state
‘n’.
 Hill-Climbing
 Often called greedy search algorithm, does
not maintain any information to backtrack to
other paths in the state space if it adopts a
fruitless “opportunist” path.
© A.I