path consistency, constraint problem solving strategies, and constraint propagation Flashcards

1
Q

Path consistency (PC)

A

How to strengthen the consistency level
require consistency over more than one constraint
-Path(V0,V1,…,Vm) is path consistent iff for every pair of values xED0 and yEDm satisfying all the binary constraints on V0,Vm there exists an assignment of variables V1,…,Vm-1 such that all the binary constraints between the neighboring variables Vi,Vi+1 are satisfied

-only guarantees paths between neighboring variables are satisfied

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

when is a CSP path consistent?

A

iff every path is consistent

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

path consistency drawbacks

A
  • memory consumption
  • Bad ratio strength/efficiency
  • modifies the constraint network
  • still not a complete technique
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

PC memory consumption

A

because PC eliminates pairs of values, we need to keep all compatible pairs extensionally, e.g. using {0,1} matrix

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

PC bad ratio strength/efficiency

A

PC removes more or same inconsistencies than AC, but the strength/efficiency ratio is much worse than for AC

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

PC constraint network modification

A
  • adds redundant arcs (constraints) and thus it changes connectivity of the constraint network
  • complicates using heuristics derived from the structure of the original constraint network
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

How to solve constraint problems

A

depth first search
consistency techniques
combination of approaches

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

CSP depth-first search

A

depth-first search

-complete but too slow (exponential) explores visibly wrong valuations

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

CSP consistency techniques

A

consistency techniques

  • usually incomplete(inconsistent values stay in domains)
  • pretty fast(polynomial)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

CSP DFS and consistency combo

A

combinations of approaches

  • combines advantages of approaches
  • label variables step by step(backtracking)
  • maintain consistency after assigning a value
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

CSP solving techniques

A
core procedure DFS
Look Back
forward checking
partial look ahead
look ahead
MAC
MCk
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

CSP core DFS

A

assign variables one by one

ensure consistency after each assignment

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

CSP Look Back

A
  • maintain consistency among already assigned variables
  • look back= look to already assigned variables
  • if the consistency test returns a conflict (+ explanation)
  • -backtrack (basic) or
  • -backjump
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

CSP forward checking:

A

prevention technique

  • removes values from future variables which are incompatible with current assignments
  • check only future variables connected to some assigned variables by some constraint
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

CSP partial look ahead

A
  • propagate the value assigned to the current variable to all future variables
  • if DAC is maintained in reverse order w.r.t. the labeling order (aka known as DAC look ahead)
  • it is not necessary to consider constraints involving past variables other than the current one
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

CSP look ahead

A

like partial look ahead but with AC instead of DAC

17
Q

CSP MAC

A

AC performed initially

maintained after each assignment

18
Q

CSP MCk

A

maintain strong-k-consistency

chronological backtracking