Artificial Intelligence &

Machine Learning

Class No-3:
State-space
representation (Contd.)
By:
Dr. Niranjan Panigrahi
CSE, PMEC, Berhampur,Odisha
 Recap…
Identification & Definition of AI
problem
Typical NP-complete/NP-hard
problems
State-space representation
Examples of Classical AI problems
State-space representation method
Water-jug problem

2
 State Vs Node
 State represents  Node represents a
physical suitable data
configuration of a structure to
given problem represent the state
of a given problem

3
 Example-2 : 8-puzzle problem
 Given: a board situation for the 8-puzzle:

1 3 8
2 7
5 4 6
 Problem: find a sequence of moves (allowed under
the rules of the 8-puzzle game) that transform this
board situation in a desired goal situation:
1 2 3
8 4
7 6 5
4
 Initial issues to solve:
 How to represent states?
 Ex.: 1 3 8 using a 3 X 3 matrix
2 7
5 4 6
 How to formulate production rules?
 Ex.:
 express how/when squares may be moved?
 Or: express how/when the blank space is moved?
 When is a rule applicable to a state? (matching)
 How to formulate when the goal criterion is satified and
how to verify that it is?
 How/which rules to activate? (control)
5
 The (implicit) search tree
 Each state-space representation defines a search
tree:
1 3 8
2 7
5 4 6

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

goal

 But this tree is only IMPLICITLY available !! 6

 Example-3: Chess playing
 Problem: develop a program that plays chess (well).

1. A way to represent board situations in an

unambiguous way:
 Ex.:

8 List:
7 (( king_black, 8, C),
6 ( knight_black, 7, B),
5 ( pawn_black, 7, G),
4 ( pawn_black, 5, F),
3 ( pawn_white, 2, H),
2 ( king_white, 1, E))
1
A B C D E F G H 7
 Chess (2):
 2. Describe the rules that represent allowed
moves:
 Ex.:
4
3
2
1
A B C D E F G H

( (pawn_white,2,x) , (blank, 3, x), (blank, 4, x) )

add( ( pawn_white, 4, x) ),
remove( (pawn_white, 2, x) )
8
 Chess (3):
 3. Provide a way to check whether a rule is
applicable to some state:
 Ex.:
List:
8
7 (( king_black, 8, C),
6 ( knight_black, 7, B),
5 ( pawn_black, 7, G),
4 ( pawn_black, 5, F),
3 ( pawn_white, 2, H),
2 ( king_white, 1, E))
1
A B C D E F G H
( (pawn_white,2,x) , (blank, 3, x), (blank, 4, x) )
add( ( pawn_white, 4, x) ),
remove( (pawn_white, 2, x) )

Matching mechanism !! 9
 Chess (4):
 4. How to specify a state in which the goal is
reached (= a winning state):
 Ex.:

win( black )  attacked( king_white ) and

no_legal_move( king_white )

+ similar definitions for:

attacked( Piece )  …
no_legal_move( Piece )  ...

10
 Chess (5):
 5. A way to verify whether a winning state is
reached.
 Ex.:

?- win( black ).
win( black )  attacked( king_white ) and

no_legal_move( king_white )

…
Need a theorem prover ( e.g. Prolog) to verify
that the state is a winning one.
11
 Chess (6).
 6. The initial state.

A program that is ABLE to play chess.

7. A mechanism that selects in each state an

appropriate rule to apply.

The control problem !

Main focus of searching

strategies!!

12
 Chess (7).
Implicit search tree
~15
Move 1

Move 2
~ (15)2

~ (15)3 Move 3

Need very efficient search techniques to find good

paths in such combinatorial trees. 13
 Issues and trade-offs:

1. Should we search through the implicit tree or

through an implicit graph?
2. Do we need an optimal solution, or just any solution?
‘optimal path problems’
3. Can we decompose states into components on which
simple rules can in an independent way?
Problem reduction or decomposability
4. Should we search forwards from the initial state,
or backwards from a goal state?

14
 1. Avoiding loops:
search in trees or graphs ?
[ 0, 0 ]

[ 4, 0 ] [ 0, 3 ]

[ 1, 3 ] [ 4, 3 ] [ 3, 0 ]

[ 1, 0 ] [ 3, 3 ]
...
Avoids generating loops: but needs to keep
track of ALL the nodes. 15
 2. Any path, versus shortest
path, versus best path:
Ex.: 8-puzzle: any or shortest path problem.
Ex.: Traveling salesperson problem:

3000
Boston SanFrancisco
250 1700 2900
3300
1450 NewYork
1200 1700
1500
Miami 1600
Dallas

 Find a sequence of cities ABCDEA such that the

total distance is MINIMAL.
Best path problem
16
 3. Problem reduction or
problem decomposition:
Ex.: Computing symbolic integrals:
 State: the integral to compute
 Rules: integration reduction rules
 Goal: all integrals have been eliminated
 x2 + 3x + sin2x . cos2x dx
AND
x2 dx 3xdx sin2x . cos2x dx

x3/3 3 xdx
((sin2x)/2)2dx (1 - cos2x).cos2x dx
... ... ... ...
AND-OR-tree search 17
 Necessary for decomposition:
independence of states:
Ex.: Blocks world problem.
 Initially: C is on A and B is on the table.
 Rules: to move any free block to another or to the table
 Goal: A is on B and B is on C.

AND-OR-tree?
C
A B
Goal: A on B and B on C

AND

C C
A B A B
Goal: A on B Goal: B on C
18
 AND

C C
A B A B
Goal: A on B Goal: B on C

C A B B
Goal: A on B C
A

A But: branches cannot be

C B combined

19
 4. Forward versus backward
reasoning:
Initial states

Goal states

 Forward reasoning (or forward chaining): from

initial states to goal states.
20
 5. Forward versus backward
reasoning:
Initial states

Goal states

 Backward reasoning (or backward chaining): from

goal states to initial states.
21