Constraint Satisfaction Problems

AIMA: Chapter 6

CIS 391 - Intro to AI 1

 Constraint Satisfaction Problems
A CSP consists of:
• Finite set of variables X1, X2, …, Xn
• Nonempty domain of possible values for each variable
D1, D2, … Dn where Di = {v1, …, vk}
• Finite set of constraints C1, C2, …, Cm
—Each constraint Ci limits the values that variables can take,
e.g., X1 ≠ X2 A state is defined as an assignment of values to
some or all variables.
 A consistent assignment does not violate the
constraints.
 Example: Sudoku

CIS 391 - Intro to AI 2

Example: Cryptarithmetic
X3 X2 X1

 Variables: F T U W R O, X1 X2 X3
 Domain: {0,1,2,3,4,5,6,7,8,9}
 Constraints:
• Alldiff (F,T,U,W,R,O)
• O + O = R + 10 · X1
• X1 + W + W = U + 10 · X2
• X2 + T + T = O + 10 · X3
• X3 = F, T ≠ 0, F ≠ 0
CIS 391 - Intro to AI 3
Constraint satisfaction problems
 An assignment is complete when every
variable is assigned a value.
 A solution to a CSP is a complete
assignment that satisfies all constraints.
 Applications:
• Map coloring
• Line Drawing Interpretation
• Scheduling problems
—Job shop scheduling
—Scheduling the Hubble Space Telescope
• Floor planning for VLSI
 Beyond our scope: CSPs that require a solution
that maximizes an objective function.
CIS 391 - Intro to AI 4
Example: Map-coloring

 Variables: WA, NT, Q, NSW, V, SA, T

 Domains: Di = {red,green,blue}
 Constraints: adjacent regions must have different colors
• e.g., WA ≠ NT
—So (WA,NT) must be in {(red,green),(red,blue),(green,red), …}

CIS 391 - Intro to AI 5

Example: Map-coloring

Solutions are complete and consistent assignments,

• e.g., WA = red, NT = green,Q = red,NSW = green,
V = red,SA = blue,T = green

CIS 391 - Intro to AI 6

Benefits of CSP
 Clean specification of many problems, generic goal,
successor function & heuristics
• Just represent problem as a CSP & solve with general package

 CSP “knows” which variables violate a constraint

• And hence where to focus the search

 CSPs: Automatically prune off all branches that

violate constraints
• (State space search could do this only by hand-building
constraints into the successor function)

CIS 391 - Intro to AI 7

CSP Representations
 Constraint graph:
• nodes are variables
• arcs are constraints

 Standard representation pattern:

• variables with values

 Constraint graph simplifies search.

• e.g. Tasmania is an independent subproblem.

 This problem: A binary CSP:

• each constraint relates two variables

CIS 391 - Intro to AI 8

Varieties of CSPs
 Discrete variables
• finite domains:
—n variables, domain size d  O(dn) complete assignments
—e.g., Boolean CSPs, includes Boolean satisfiability (NP-complete)
—Line Drawing Interpretation
• infinite domains:
—integers, strings, etc.
—e.g., job scheduling, variables are start/end days for each job
—need a constraint language, e.g., StartJob1 + 5 ≤ StartJob3

 Continuous variables
• e.g., start/end times for Hubble Space Telescope observations
• linear constraints solvable in polynomial time by linear
programming

CIS 391 - Intro to AI 9

Varieties of constraints
 Unary constraints involve a single variable,
• e.g., SA ≠ green

 Binary constraints involve pairs of variables,

• e.g., SA ≠ WA

 Higher-order constraints involve 3 or more variables

• e.g., crypt-arithmetic column constraints

 Preference (soft constraints) e.g. red is better than

green can be represented by a cost for each variable
assignment
• Constrained optimization problems.

CIS 391 - Intro to AI 10

Idea 1: CSP as a search problem

 A CSP can easily be expressed as a search

problem
• Initial State: the empty assignment {}.
• Successor function: Assign value to any unassigned
variable provided that there is not a constraint conflict.
• Goal test: the current assignment is complete.
• Path cost: a constant cost for every step.

 Solution is always found at depth n, for n variables

• Hence Depth First Search can be used

CIS 391 - Intro to AI 11

Backtracking search
 Note that variable assignments are commutative
• Eg [ step 1: WA = red; step 2: NT = green ]
equivalent to [ step 1: NT = green; step 2: WA = red ]
• Therefore, a tree search, not a graph search

 Only need to consider assignments to a single variable at

each node
• b = d and there are dn leaves

 Depth-first search for CSPs with single-variable

assignments is called backtracking search

 Backtracking search is the basic uninformed algorithm for

CSPs

 Can solve n-queens for n ≈ 25

CIS 391 - Intro to AI 12

Backtracking example

And so on….

CIS 391 - Intro to AI 14

Backtracking example

And so on….

CIS 391 - Intro to AI 15

Idea 2: Improving backtracking efficiency
 General-purpose methods & heuristics can give
huge gains in speed, on average
 Heuristics:
• Q: Which variable should be assigned next?
1. Most constrained variable
2. Most constraining variable

• Q: In what order should that variable’s values be tried?

3. Least constraining value

• Q: Can we detect inevitable failure early?

4. Forward checking

CIS 391 - Intro to AI 16

Heuristic 1: Most constrained variable
 Choose a variable with the fewest legal values

 a.k.a. minimum remaining values (MRV) heuristic

CIS 391 - Intro to AI 17

Heuristic 2: Most constraining variable
 Tie-breaker among most constrained variables
 Choose the variable with the most constraints
on remaining variables

CIS 391 - Intro to AI 18

Heuristic 3: Least constraining value
 Given a variable, choose the least constraining
value:
• the one that rules out the fewest values in the remaining
variables

 Combining these heuristics makes 1000 queens

feasible

Note: demonstrated here independent of the

other heuristics
CIS 391 - Intro to AI 19
Heuristic 4: Forward checking
 Idea:
• Keep track of remaining legal values for unassigned
variables
• Terminate search when any variable has no legal values,
given its neighbors

• (For later: Edge & Arc consistency are variants)

CIS 391 - Intro to AI 20

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

CIS 391 - Intro to AI 21

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

CIS 391 - Intro to AI 22

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

CIS 391 - Intro to AI 23

Example: 4-Queens Problem

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

(From Bonnie Dorr, U of Md, CMSC 421)

CIS 391 - Intro to AI 24

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}

CIS 391 - Intro to AI 25

Example: 4-Queens Problem

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

CIS 391 - Intro to AI 26

Example: 4-Queens Problem

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

CIS 391 - Intro to AI 27

Example: 4-Queens Problem

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

CIS 391 - Intro to AI 28

 Example: 4-Queens Problem

Picking up a little later after two

steps of backtracking….
X1 X2
1 2 3 4 { ,2,3,4} {1,2,3,4}
1
2
3
4
X3 X4
{1,2,3,4} {1,2,3,4}

CIS 391 - Intro to AI 29

Example: 4-Queens Problem

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

CIS 391 - Intro to AI 30

Example: 4-Queens Problem

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

CIS 391 - Intro to AI 31

Example: 4-Queens Problem

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

CIS 391 - Intro to AI 32

Example: 4-Queens Problem

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

CIS 391 - Intro to AI 33

Example: 4-Queens Problem

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

CIS 391 - Intro to AI 34

Example: 4-Queens Problem

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

CIS 391 - Intro to AI 35

Towards Constraint propagation
 Forward checking propagates information from
assigned to unassigned variables, but doesn't
provide early detection for all failures:

 NT and SA cannot both be blue!

 Constraint propagation goes beyond forward
checking & repeatedly enforces constraints locally

CIS 391 - Intro to AI 36

 Interpreting line drawings and the invention
of constraint propagation algorithms

CIS 391 - Intro to AI 37

We Interpret Line Drawings As 3D
 We have strong intuitions about line drawings of
simple geometric figures:
 We naturally interpret 2D line drawings as planar
representations of 3D objects.
 We interpret each line as being either a convex,
concave or occluding edge in the actual object.

CIS 391 - Intro to AI 38

Interpretation as Convexity Labeling
Each edge in an image can be interpreted to be either a convex
edge, a concave edge or an occluding edge:

 + labels a convex edge (angled toward the viewer);

 - labels a concave edge (angled away from the viewer);
 labels an occluding edge. To its right is the body for
which the arrow line provides an edge. On its left is space.

occluding
convex concave

CIS 391 - Intro to AI 39

Huffman/Clowes
Line Drawing Interpretation
 Given: a line drawing of a simple “blocks
world” physical image
 Compute: a set of junction labels that yields
a consistent physical interpretation

CIS 391 - Intro to AI 40

Huffman/Clowes Junction Labels
 A simple trihedral image can be automatically
interpreted given only information about each
junction in the image.
 Each interpretation gives convexity information for
each junction.
 This interpretation is based on the junction type. (All
junctions involve at most three lines.)

arrow Y L T
junction junction junction junction

CIS 391 - Intro to AI 41

Arrow Junctions have only 3 interpretations

A2

A3
A1

The image of the same vertex from a different point of

view gives a different junction type
(from Winston, Intro to Artificial Intelligence)
CIS 391 - Intro to AI 42
L Junctions have only 6 interpretations!
L4
L2

L1
L3

L5 L6

CIS 391 - Intro to AI 43

The world constrains possibilities

Type of Physically Possible Combinatoric

Junction Interpretations Possibilities

Arrow 3 4x4x4= 64

L 6 4x4=16

T 4 4x4x4=64

Y 3 4x4x4=64

CIS 391 - Intro to AI 44

Idea 3 (big idea): Inference in CSPs
 CSP solvers combine search and inference
• Search
—assigning a value to a variable
• Constraint propagation (inference)
—Eliminates possible values for a variable
if the value would violate local consistency
• Can do inference first, or intertwine it with search
—You’ll investigate this in the Sudoku homework
 Local consistency
• Node consistency: satisfies unary constraints
—This is trivial!
• Arc consistency: satisfies binary constraints
—Xi is arc-consistent w.r.t. Xj if for every value v in Di, there is some
value w in Dj that satisfies the binary constraint on the arc between
Xi and Xj.
CIS 391 - Intro to AI 45
 An Example Constraint:
The Edge Consistency Constraint

CIS 391 - Intro to AI 46

The Edge Consistency Constraint

Any consistent assignment of labels to the

junctions in a picture must assign the
same line label to any given line.

...
...
+
+ L3 ...
L3 ... +
-
L4
L5

CIS 391 - Intro to AI 47

An Example of Edge Consistency
 Consider an arrow junction with an L junction to the right:

 A1 and either L1 or L6 are consistent since they both

associate the same kind of arrow with the line.

 A1 and L2 are inconsistent, since the arrows are pointed in

the opposing directions,

 Similarly, A1 and L3 are inconsistent.

CIS 391 - Intro to AI 48

Replacing Search:
Constraint Propagation Invented…
Dave Waltz’s insight for line labeling:
 Pairs of adjacent junctions (junctions connected by
a line) constrain each other’s interpretations!

 By iterating over the graph, the edge-consistency

constraints can be propagated along the connected
edges of the graph.

 Search: Use constraints to add labels to find one

solution
 Constraint Propagation: Use constraints to
eliminate labels to simultaneously find all solutions

CIS 391 - Intro to AI 49

The Waltz/Mackworth Constraint
Propagation Algorithm for line labeling
1. Assign every junction in the picture a set of all
Huffman/Clowes junction labels for that junction
type;
2. Repeat until there is no change in the set of
labels associate with any junction:
3. For each junction i in the picture:
4. For each neighboring junction j in the picture:
5. Remove any junction label from i for which
there is no edge-consistent junction label
on j.

CIS 391 - Intro to AI 50

 Waltz/Mackworth: An example
A1,A2, A1,A2,
A3 A3 A1,A3

Check

L1,L2,L3, Given
L1,L3, A1,A2, L1,L3, A1,A2,
L4,L5, A1,A2, L5,L6 L5,L6
L6 A3 A3 A3
A1,A3 A1,A3

…
L1, A1,A2, L1, A1, A3
L5,L6 A3 L5,L6
CIS 391 - Intro to AI 51
Inefficiencies: Towards AC-3
1. At each iteration, we only need to examine those
Xi where at least one neighbor of Xi has lost a value
in the previous iteration.
2. If Xi loses a value only because of edge
inconsistencies with Xj, we don’t need to check Xj
on the next iteration.
3. Removing a value on Xi can only make Xj edge-
inconsistent is with respect to Xi itself. Thus, we
only need to check that the labels on the pair (j,i) are
still consistent.
These insights lead a much better algorithm...

CIS 391 - Intro to AI 52

Directed arcs, Not undirected edges
 Given a pair of nodes Xi and Xj in the constraint
graph connected by a constraint edge, we
represent this not by a single undirected edge,
but a pair of directed arcs.
 For a connected pair of junctions Xi and Xj , there
are two arcs that connect them: (i,j) and (j,i).

(i,j)

Xi Xj

CIS 391 - Intro to AI (j,i) 53

Arc consistency: the general case
 Simplest form of propagation makes each arc
consistent
 X Y is consistent iff
for every value x of X there is some allowed y in Y

CIS 391 - Intro to AI 54

Arc Consistency
An arc (i,j) is arc consistent if and only if every value v on Xi is
consistent with some label on Xj.

 To make an arc (i,j) arc consistent,

for each value v on Xi ,
if there is no label on Xj consistent with v then
remove v from Xi

CIS 391 - Intro to AI 55

When to Iterate, When to Stop?
The crucial principle:
If a value is removed from a node Xi,
then the values on all of Xi’s neighbors must be
reexamined.

Why? Removing a value from a node may result in

one of the neighbors becoming arc inconsistent,
so we need to check…

(but each neighbor Xj can only become inconsistent

with respect to the removed values on Xi)

CIS 391 - Intro to AI 56

AC-3
function AC-3(csp) return the CSP, possibly with reduced domains
inputs: csp, a binary csp with variables {X1, X2, …, Xn}
local variables: queue, a queue of arcs initially the arcs in csp
while queue is not empty do
(Xi, Xj)  REMOVE-FIRST(queue)
if REMOVE-INCONSISTENT-VALUES(Xi, Xj) then
for each Xk in NEIGHBORS[Xi ] – {Xj} do
add (Xk, Xi) to queue

function REMOVE-INCONSISTENT-VALUES(Xi, Xj) return true iff we remove a value

removed  false
for each x in DOMAIN[Xi] do
if no value y in DOMAIN[Xj] allows (x,y) to satisfy the constraints between
Xi and Xj
then delete x from DOMAIN[Xi]; removed  true
return removed

CIS 391 - Intro to AI 57

AC-3: Worst Case Complexity Analysis
 All nodes can be connected to every other node,
• so each of n nodes must be compared against n-1 other nodes,
• so total # of arcs is 2*n*(n-1), i.e. O(n2)
 If there are d values, checking arc (i,j) takes O(d2) time
 Each arc (i,j) can only be inserted into the queue d times
 Worst case complexity: O(n2d3)

For planar constraint graphs, the number of arcs can only be

linear in N, so for our pictures, the time complexity is only
O(nd3)
• The constraint graph for line drawings is isomorphic to the line
drawing itself, so is a planar graph.

CIS 391 - Intro to AI 58

Beyond binary constraints:
Path consistency
• Generalizes arc-consistency from individual binary
constraints to multiple constraints
• A pair of variables Xi, Xj is path-consistent w.r.t. Xm if for
every assignment Xi=a, Xj=b consistent with the constraints
on Xi, Xj there is an assignment to Xm that satisfied the
constraints on Xi, Xm and Xj, Xm
 Global constraints
• Can apply to any number of variables
• E.g., in Sudoko, all numbers in a row must be different
• E.g., in cryptarithmetic, each letter must be a different digit
• Example algorithm:
—If any variable has a single possible value, delete that variable
from the domains of all other constrained variables
—If no values are left for any variable, you found a contradiction

CIS 391 - Intro to AI 59

Chronological backtracking
 DFS does Chronological backtracking
• If a branch of a search fails, backtrack to the most recent
variable assignment and try something different
• But this variable may not be related to the failure
 Example: Map coloring of Australia
• Variable order
—Q, NSW, V, T, SA, WA, NT.
• Current assignment:
—Q=red, NWS=green, V=blue, T= red
• SA cannot be assigned anything
• But reassigning T does not help!

CIS 391 - Intro to AI 60

Backjumping: Improved backtracking
 Find “the conflict set”
• Those variable assignments that are in conflict
• Conflict set for SA: {Q=red, NSW=green, V=blue}
 Jump back to reassign one of those conflicting
variables
 Forward checking can build the conflict set
• When a value is deleted from a variable’s domain, add it to
its conflict set
• But backjumping finds the same conflicts that forward
checking does
• Fix using “conflict-directed backjumping”
—Go back to predecessors of conflict set

CIS 391 - Intro to AI 61

Local search for CSPs
 Hill-climbing, simulated annealing typically work
with "complete" states, i.e., all variables assigned

 To apply to CSPs:
• allow states with unsatisfied constraints
• operators reassign variable values

 Variable selection: randomly select any conflicted

variable

 Value selection by min-conflicts heuristic:

• choose value that violates the fewest constraints
• i.e., hill-climb with h(n) = total number of violated constraints

CIS 391 - Intro to AI 62

Example: n-queens
 States: 4 queens in 4 columns (44 = 256 states)
 Actions: move queen in column
 Goal test: no attacks
 Evaluation: h(n) = number of attacks

 Given random initial state, local min-conflicts can

solve n-queens in almost constant time for
arbitrary n with high probability (e.g., n =
10,000,000)

CIS 391 - Intro to AI 63

Simple CSPs can be solved quickly
1. Completely independent subproblems
• e.g. Australia & Tasmania
• Easiest
2. Constraint graph is a tree
• Any two variables are connected by only a single path
• Permits solution in time linear in number of variables
• Do a topological sort and just march down the list
A E
\ /
B– D A  B  C  D E  F
/ \
C F

CIS 391 - Intro to AI 64

Simplifying hard CSPs: Cycle Cutsets
 Constraint graph can be decomposed into a tree
• Collapse or remove nodes
• Cycle cutset S of a graph G: any subset of vertices of G that, if
removed, leaves G a tree
 Cycle cutset algorithm
• Choose some cutset S
• For each possible assignment to the variables in S that
satisfies all constraints on S
—Remove any values for the domains of the remaining variables
that are not consistent with S
—If the remaining CSP has a solution, then you have are done
• For graph size n, domain size d
—Time complexity for cycle cutset of size c:
O(dc * d2(n-c)) = O(dc+2 (n-c))

CIS 391 - Intro to AI 65