CS440 Lectures https://courses.grainger.illinois.edu/cs440/fa2020/lectures/csp...

CS 440/ECE 448
Fall 2020 Constraint Satisfaction Problems 1
Margaret Fleck

The last few topics in this course cover specialized types of search. These were once major
components of AI systems. They are are still worth knowing about because they still appear
in the higher-level (more symbolic) parts of AI systems. We'll start by looking at constraint
satisfaction problems.

The n-queens problem

Remember that a queen in chess threatens another piece if there is a direct path between
the two either horizontally, or vertically, or diagonally. In the n queens problem, we have
to place n queens on a n by n board so that no queen threatens another. Here is a solution
for n=8:

From Wikipedia

Since each column must contain exactly one queen, we can model this as follows

One variable per column.

Possible values: the n vertical positions within the column.
Constraints: can't have two in the same row, can't have two in the same diagonal.

We can visualize this as a graph in which each variable is a node and edges link variables
involved in binary constraints. So X1 in the following diagram is a variable representing a
column. Its possible values {1,2,3,4} are the row to place the queen in.

1 of 5 5/10/21, 01:48
CS440 Lectures https://courses.grainger.illinois.edu/cs440/fa2020/lectures/csp...

from Bonnie Dorr, U. Maryland (via Mitch Marcus at U. Penn).

What is a CSP problem?

A constraint-satisfaction problem (CSP) consists of

Variables
Set of allowed values (for each variable)
Constraints

In many examples, all variables have the same set of allowed values. But there are problems
where different variables may have different possible values.

Need to find a complete assignment (one value for each variable) that satisfies all the
constraints

In basic search problems, we are explicitly given a goal (or perhaps a small set of goals) and
need to find a path to it. In a constraint-satisfaction problem, we're given only a description
of what constraints a goal state must satisfy. Our problem is to find a goal state that satisfies
these constraints. We do end up finding a path to this goal, but the path is only temporary
scaffolding.

Map Coloring
The map coloring problem requires that we color regions of the plan such that neighboring
regions never have the same color. (Touching only at a corner doesn't count as neighboring.)
The map below shows an example.

2 of 5 5/10/21, 01:48
CS440 Lectures https://courses.grainger.illinois.edu/cs440/fa2020/lectures/csp...

From Wikipedia

Viewed as a CSP problem

Variables: states/countries
Values: {pink, orange, green, yellow}
Constraints: extended boundary --> different color

Pictured as a constraint graph, the states/countries are nodes, and edges connect adjacent
states. A small example is shown below. Theoretical work on this topic is found mostly
under the keyword "graph coloring." Applications of graph coloring include register
allocation, final exam scheduling.

The "4 color Theorem" states that any planar map can be colored with only four colors. It
was proved at U. Illinois in 1976 by Kenneth Appel and Wolfgang Haken. It required a
computer program to check almost 2000 special cases. This computer-assisted proof was
controversial at the time, but became accepted after the computerized search had been
replicated a couple times.

Notice that graph coloring is NP complete. We don't know for sure if NP problems require
polynomial or exponential time, but we suspect they require exponential time. However,
many practical applications can exploit domain-specific heuristics (e.g. linear scan for
register allocation) or loopholes (e.g. ok to have small conflicts in final exams) to produce
fast approximate algorithms.

3 of 5 5/10/21, 01:48
CS440 Lectures https://courses.grainger.illinois.edu/cs440/fa2020/lectures/csp...

Cryptarithmetic
Cryptarithmetic is another classic CSP problem. Here's an example:

T W O
+ T W O
---------
= F O U R

We can make this into a CSP model as follows:

Variables: the letters

Possible values: {0,1,..,9}
Constraints: all letters have different values, addition works as intended, leading digits
aren't zero.

"Addition works as intended" can be made concrete with these equations, where C_1 and
C_10 are the carry values.

O + O = R + 10 * C_1
W + W + C_1 = U + 10*C_10
T + T + C_10 = O + 10*F

"All different" and the equational constraints involve multiple variables. Usually best to
keep them in that form, rather than converting to binary. We can visualize this using a
slightly different type of graph with some extra nodes:

nodes for the basic variables

nodes for auxiliary variables (carry values)
square boxes for constraints, attached to variables they involve

Sudoku

4 of 5 5/10/21, 01:48
CS440 Lectures https://courses.grainger.illinois.edu/cs440/fa2020/lectures/csp...

from Wikipedia

Sudoku can be modelled as a CSP problem,

Variables: cells x_ij

Values: {1,2,...,9}
Constraints: no duplicates in row, no duplicates in column, no duplicates in block

Scheduling problems
We can also use this framework to model scheduling problems, e.g. planning timing and
sequence of operations in factory assembly.

Variables: tasks (with how long they take)

Values: start times
Constraints: certain tasks must/must not run at the same time, certain tasks must
precede/follow other tasks.

5 of 5 5/10/21, 01:48