State Space Representation 3.
0
of Problems
3.1 Graphs
3.2 Formulating Search Problems
3.3 The 8-Puzzle as an example
3.4 State Space Representation using graphs
3.5 Performing a State Space Search
3.6 Basic Depth First Search (DFS)
3.7 Basic Breadth First Search (DFS)
3.8 Best First Search (DFS)
3.1 Graphs
Definitions:
a graph consists of:
3
4
A set of nodes N1, N2, N3,Nn.
A set of arcs that connect pairs
of nodes.
A directed graph has an
indicated direction for
traversing each arc.
A labeled graph has its
nodes labeled.
A labeled directed graph is
shown in figure 4.2
5
Figure 4.1:5 nodes, and 6
arcs graph.
a
b
c
d
e
Figure 4.2: Labeled directed graph
Nodes {a,b,c,d,e}
Arcs:{(a,b),(b,e),(c,a),(c,b), (d,c),
(e,d)}
A path through a graph connects a
sequence of nodes through
successive arcs. It is represented
by an ordered list of the nodes
representing the path.
For example in figure 4.3, [a, b, e,
d] is a path through nodes a ,b ,e , d.
A rooted graph has a unique node
(called the root ) such that there is
a path from the root to all nodes
within the graph. i.e. all paths
originate from the root ( figure
4.4).
a
b
c
d
e
Figure 4.3: dotted curve
indicates the path
[a,b,e,d]
The root
Figure 4.4: a rooted
graph
A tree is a graph in which
each two nodes have at
most one path between
them.
Figure 4.5 is an example of a
rooted tree.
If a directed arc connects Ni
to Nk then
Ni is the parent of Nk and
Nk is the child of Ni..
In figure 4.5: d is the parent
of e and f.
e and f are called siblings.
The root
b
h
Figure 4.5: a rooted tree
In a graph:
1. An ordered sequence of nodes [ N1, N2, N3 ..,
Nn], where each Ni, Ni+1 in the sequence
represent an arc (Ni,Ni+1), is called a path of
length n-1.
2. If a path contains any node more than once it
said to contain a cycle or loop.
3. Two nodes in a graph are said to be
connected if there is a path that includes
them both.
4. On a path on a rooted graph, a node is said
to be the ancestor of all nodes positioned
after it ( to its right) as well as descendent of
all nodes before it ( to its left)
-For example, in figure 4.5, d is the ancestor of e,
while it is the descendent of a in the path [a, d, e].
3.2 Formulating Search Problems
All search problems can be cast into the
following general form:
Starting State
E.g.
starting city for a route
Goal State (or a test for goal state)
E.g.
destination city
The permissible operators
E.g.
go to city X
A state is a data structure which captures all
relevant information about the problem.
E.g.
a node on a partial path
3.3 The 8-Puzzle as an example
The eight puzzle consists of a 3 x 3 grid with 8
consecutively numbered tiles arranged on it. Any
tile adjacent to the space can be moved on it. A
number of different goal states are used.
A state for this problem needs to keep
track of the position of all tiles on the
game board, with 0 representing the blank
position (space) on the board
The initial state could be represented as:
( (5,4,0), (6,1,8), (7,3,2) )
The final state could be represented as:
( (1,2,3) (8,0,4), (7,6,5) )
The operators can be thought of in terms
of the direction that the blank space
effectively moves. i.e.. up, down, left, right.
5 4 .
1 2 3
6 1 8
8 . 4
7 3 2
7 6 5
Start State Goal State
State Space Representation Using Graphs 3.4
In the state space representation of a
problem:
nodes of a graph correspond to partial
problem solution states.
arcs correspond to steps (application of
operators) in a problem solving process.
The root of the graph corresponds to the
initial state of the problem.
The goal node which may not exist, is a leaf
node which corresponds to a goal state.
State Space Search is the process of
finding a solution path from the start state
to a goal state.
The task of a search algorithm is to find a
solution path through such a problem
space.
The generation of new states ( expansion
of nodes) along the path is done by
applying the operators (such as legal
moves in a game).
A goal may describe
a statea winning board in a simple game.
or some property of the solution path itself
(length of the path) shortest path for example.
3.5 Performing a State Space Search
State space search involves finding a path from the initial state
of a search problem to a goal state.
To do this,
1-build a search graph, starting from the initial state (or the
goal state)
2- expand a state by applying the search operators to that
state, generating ALL of its successor states.
These successors are in the next level down of the search graph
3-The order in which we choose states for expansion is
determined by the search strategy
Different strategies result in (sometimes massively) different
behaviour
KEY CONCEPT: We want to find the solution while realizing
in memory as few as possible of the nodes in the search space.
3.6 Basic Depth First Search (DFS)
/* OPEN and CLOSED are lists */
OPEN = Start node, CLOSED = empty
While OPEN is not empty do
Remove leftmost state from OPEN, call it X
If X is a goal return success
Put X on CLOSED
Generate all successors of X
Eliminate any successors that are already
on OPEN or CLOSED
put remaining successors
on LEFT end of OPEN
End while
Note:
For depth first put successors on LEFT (i.e. acts like a STACK)
For breadth first put successors on Right (i.e. acts like a QUEUE)
Consider the following segment of a search for a solution to
the 8-Puzzle problem
a.The initial state
b. After expanding that state
c. After expanding "last" successor
generated
In depth first search, the "last" successor generated will
. be expanded next
Example: road map
Consider the following road map
A
Applying the DFS
1-Open=[S],
closed=[]
2-Open=[AC], closed=[S]
3-Open=[BC], closed=[AS]
4-Open=[DC], closed=[BAS]
A
5-Open=[GC], closed=[DBAS]
6-Open=[C],
closed=[GDBAS]
Report success
Path is SABDG
D
A
B
D
G
3.7 Basic Breadth First Search (BRFS)
/* OPEN and CLOSED are lists */
OPEN = Start node, CLOSED = empty
While OPEN is not empty do
Remove leftmost state from OPEN, call it X
If X is a goal return success
Put X on CLOSED
Generate all successors of X
Eliminate any successors that are already
on OPEN or CLOSED
put remaining successors
on RIGHT end of OPEN
End while
Note:
For depth first put successors on LEFT (i.e. acts like a STACK)
For breadth first put successors on RIGHT (i.e. acts like a QUEUE)
Example: road map
Consider the following road map
A
Applying the BRFS
1-Open=[S],
closed=[]
2-Open=[AC], closed=[S]
3-Open=[CB], closed=[AS]
4-Open=[BD], closed=[CAS]
A
5-Open=[D], closed=[CAS]
6-Open=[G],
closed=[DCAS]
7-Open=[],
closed=[GDCAS]
Report success
Path is SCDG
D
G
3.8 Best First Search (BFS)
/* OPEN and CLOSED are lists */
OPEN = Start node, CLOSED = empty
While OPEN is not empty do
Remove leftmost state from OPEN, call it X
If X is a goal return success
Put X on CLOSED
Generate all successors of X
Eliminate any successors that are already
on OPEN or CLOSED
put remaining successors on OPEN sorted
according to their heuristic distance to the
goal
( ascending from left to right)
End while
Example: road map
Consider the following road map
A
Applying the BFS
1-Open=[S],
closed=[]
3
S
2-Open=[C7A8], closed=[S]
3-Open=[D3A8], closed=[CS]
4-Open=[G0B3A8], closed=[DCS]
A
5-Open=[B3A8], closed=[GDCS]
Report success
Path is SCDG
A
B
D
G
