Constraint Satisfaction Problem-II

LECTURE # 08
TUESDAY, FEBRUARY 14, 2017
SPRING 2016
FAST – NUCES, FAISALABAD CAMPUS

Zain Iqbal
Zain.iqbal@nu.edu.pk
Agenda
 Job-Shop Scheduling
 Forward Checking
 Constraint Propagation
• Node consistency
• Arc consistency
• Path consistency
• K-consistency
• Global contraints
 Intelligent Backtracking
• Constraint Learning
Job-Shop Scheduling
 Small part of the car assembly, consisting of 15 tasks
 X = {AxleF , AxleB, WheelRF ,WheelLF ,WheelRB,WheelLB,
NutsRF ,NutsLF , NutsRB, NutsLB, CapRF , CapLF , CapRB,
CapLB, Inspect} .
 Precedence constraints
• AxleF + 10 ≤ WheelRF ; AxleF + 10 ≤ WheelLF
• AxleB + 10 ≤ WheelRB; AxleB +10 ≤ WheelLB
• WheelRF + 1 ≤ NutsRF ; NutsRF + 2 ≤ CapRF ;
• WheelLF + 1 ≤ NutsLF ; NutsLF +2 ≤ CapLF ;
• WheelRB + 1 ≤ NutsRB; NutsRB + 2 ≤ CapRB;
• WheelLB + 1 ≤ NutsLB; NutsLB + 2 ≤ CapLB
 Inspect we add a constraint of the form X +dX ≤ Inspect
 Done in 30 min So Di = {1, 2, 3, . . . , 27} .
3
Heuristic 4: Forward checking
• Idea:
• Keep track of remaining legal values for unassigned
variables
• Terminate search when any variable has no legal values

• Edge & Arc consistency are variants

4
Forward checking
• Idea:
• Keep track of remaining legal values for unassigned
variables
• Terminate search when any variable has no legal
values

5
Forward checking
• Idea:
• Keep track of remaining legal values for unassigned
variables
• Terminate search when any variable has no legal values

6
Forward checking
• Idea:
• Keep track of remaining legal values for unassigned variables
• Terminate search when any variable has no legal values

• A Step toward AC-3: The most efficient algorithm

7
Example: 4-Queens Problem

X1 X2
{1,2,3,4} {1,2,3,4}
1 2 3 4
1
2
3
4 X3 X4
{1,2,3,4} {1,2,3,4}

(From B.J. Dorr, U of Md, CMSC 421)

8
Example: 4-Queens Problem

X1 X2
{1,2,3,4} {1,2,3,4}
1 2 3 4
1
2
3
4 X3 X4
{1,2,3,4} {1,2,3,4}

9
Example: 4-Queens Problem

X1 X2
{1,2,3,4} { , ,3,4}
1 2 3 4
1
2
3
4 X3 X4
{ ,2, ,4} { ,2,3, }

10
Example: 4-Queens Problem

X1 X2
{1,2,3,4} { , ,3,4}
1 2 3 4
1
2
3
4 X3 X4
{ ,2, ,4} { ,2,3, }

11
Example: 4-Queens Problem

X1 X2
{1,2,3,4} { , ,3,4}
1 2 3 4
1
2
3
4 X3 X4
Backtrack!!!
{ , , , } { ,2,3, }

12
Example: 4-Queens Problem

X1 X2
{ ,2,3,4} {1,2,3,4}
1 2 3 4
1
2
3
4 X3 X4
{1,2,3,4} {1,2,3,4}

13
Example: 4-Queens Problem

X1 X2
{ ,2,3,4} { , , ,4}
1 2 3 4
1
2
3
4 X3 X4
{1, ,3, } {1, ,3,4}

14
Example: 4-Queens Problem

X1 X2
{ ,2,3,4} { , , ,4}
1 2 3 4
1
2
3
4 X3 X4
{1, ,3, } {1, ,3,4}

15
Example: 4-Queens Problem

X1 X2
{ ,2,3,4} { , , ,4}
1 2 3 4
1
2
3
4 X3 X4
{1, , , } {1, ,3, }

16
Example: 4-Queens Problem

X1 X2
{ ,2,3,4} { , , ,4}
1 2 3 4
1
2
3
4 X3 X4
{1, , , } {1, ,3, }

17
Example: 4-Queens Problem

X1 X2
{ ,2,3,4} { , , ,4}
1 2 3 4
1
2
3
4 X3 X4
{1, , , } { , ,3, }

18
Example: 4-Queens Problem

X1 X2
{ ,2,3,4} { , , ,4}
1 2 3 4
1
2
3
4 X3 X4
{1, , , } { , ,3, }

19
20
Constraint propagation
• Forward checking only looks at variables connected to current value in
constraint graph.

• NT and SA cannot both be blue!

• Constraint propagation repeatedly enforces constraints locally
• An algorithm can search (choose a new variable assignment from
several possibilities) or do a specific type of inference called constraint
propagation

21
Node Consistency

22
Arc consistency
• Simplest form of propagation makes each arc consistent
• X Y is consistent iff
for every value x of X there is some allowed y

consistent arc.

constraint propagation propagates arc consistency on the graph.

23
Arc consistency
• Simplest form of propagation makes each arc consistent
• X Y is consistent iff
for every value x of X there is some allowed y

inconsistent arc.
remove blue from source consistent arc.

24
Arc consistency

• Simplest form of propagation makes each arc consistent

• X Y is consistent iff
for every value x of X there is some allowed y

this arc just became inconsistent

• If X loses a value, neighbors of X need to be rechecked:

i.e. incoming arcs can become inconsistent again
(outgoing arcs will stay consistent).

25
 Arc consistency
• Simplest form of propagation makes each arc consistent
• X Y is consistent iff
for every value x of X there is some allowed y

• If X loses a value, neighbors of X need to be rechecked

• Arc consistency detects failure earlier than forward checking
• Can be run as a preprocessor or after each assignment

 Time complexity: O(n2d3) 26

 Arc Consistency
• This is a propagation algorithm. It’s like sending messages to neighbors on the
graph! How do we schedule these messages?

• Every time a domain changes, all incoming messages need to be re-send.

Repeat until convergence  no message will change any domains.

• Since we only remove values from domains when they can never be part of a
solution, an empty domain means no solution possible at all  back out of that
branch.

• Forward checking is simply sending messages into a variable that just got its
value assigned. First step of arc-consistency.

27
AC-3 Algorithm

Time complexity: O(n2d3)

 From there, AC-3 does constraint propagation in the usual way, and
if any variable has its domain reduced to the empty set, the call to
AC-3 fails and we know to backtrack immediately.
28
 Path Consistency
Arc consistency fail for only two constraints
• Path consistency tightens the binary constraints by using
implicit constraints that are inferred by looking at triples of
variables
 A two-variable set {Xi,Xj} is path-consistent with respect to a third variable Xm
if, for every assignment {Xi = a,Xj = b} consistent with the constraints on
{Xi,Xj}, there is an assignment to Xm that satisfies the constraints on {Xi,Xm}
and {Xm,Xj}. This is called path consistency because one can think of it as
looking at a path from Xi to Xj with Xm in the middle.
• If There are only two: {WA = red ,SA = blue} and {WA = blue,
SA = red}.
• We can see that with both of these assignments NT can be
neither red nor blue
(because it would conflict with either WA or SA).
• Because there is no valid choice for NT

29
K-consistency
• Arc consistency does not detect all inconsistencies:
Partial assignment {WA=red, NSW=red} is
inconsistent.
• Stronger forms of propagation can be defined using
the notion of k-consistency.
• A CSP is k-consistent if for any set of k-1 variables
and for any consistent assignment to those
variables, a consistent value can always be assigned
to any kth variable.
E.g. 1-consistency or node-consistency
E.g. 2-consistency or arc-consistency
E.g. 3-consistency or path-consistency
30
Strongly k-consistent
• A graph is strongly k-consistent if
It is k-consistent and
Is also (k-1) consistent, (k-2) consistent, … all the way
down to 1-consistent.
• This is ideal since a solution can be found in time
instead of
• YET no free lunch: any algorithm for establishing
n-consistency must take time exponential in n, in
the worst case it will require exponential space n.

31
 Further improvements
Checking special constraints
Checking Alldif(…) constraint
• E.g. {WA=red, NSW=red}
 Checking Atmost(…) constraint
• E.g. Atmost(10, P1, P2, P3, P4).
• Domain {2, 3, 4, 5, 6}, the values 5 and 6 can be deleted
• Checking bounds propagation(…)
Bounds propagation for larger value domains
Bounds propagation for Smaller value domains
• E.g. Flight Passenger D1 = [0, 165] and D2 = [0, 385] .
OR
• F1 + F2 = 420. D1 = [35, 165] and D2 = [255, 385] .

32
Intelligent backtracking:
looking backward

33
Intelligent backtracking
Standard form is chronological backtracking i.e. try
different value for preceding variable.
 More intelligent, backtrack to conflict set.
• Set of variables that caused the failure or set of previously
assigned variables that are connected to X by constraints.
• Back-jumping moves back to most recent element of the
conflict set.
• Forward checking can be used to determine conflict set.

34
 Example-Map Coloring

When we try the next variable, SA, we see that every value violates a constraint. We
back up to T and try a new color for Tasmania!-But its Silly

• we will keep track of a set of assignments that are in conflict

with some value for SA.
• The set in this case {Q=red ,NSW =green, V =blue, }, is called
the conflict set for SA

 The back-jumping method backtracks to the most

recent assignment in the conflict set; in this case,
back-jumping would jump over Tasmania and try
a new value for V .

Why Back-Jumping Over Forward Checking?

35
 Example-Map Coloring
• Consider again the partial assignment {WA=red ,NSW
=red}
 Which we know is inconsistent
• Suppose we try T =red next and then assign NT, Q, V , SA.
• We know that no assignment can work for these last four
variables, so eventually we run out of values to try at NT.

Where to backtrack?

 The four variables NT, Q, V , and SA, taken together,

failed because of a set of preceding variables, which
must be those variables that directly conflict with
the four.
 In this case, the set is WA and NSW, so the
algorithm should backtrack to NSW and skip
over Tasmania.
Conflict-directed back-jumping 36
Constraint learning
• Constraint learning is the idea of finding a minimum set of
variables from the conflict set that causes the problem
• This set of variables, along with their corresponding values,
is called a no-good.
• We then record the no-good, either by adding a new
constraint to the CSP or by keeping a separate cache of no-
goods.
• No-goods can be effectively used by forward checking or by
back-jumping.
• Constraint learning is one of the most important techniques
used by modern CSP solvers to achieve efficiency on
complex problems.

37
Reading Material
• Russell & Norvig: Chapter # 7
• David Poole: Chapter # 4

38