3 - Informed Search Flashcards

1
Q

Informed Search Techniques …

A

… have additional information about more promising paths and are therefore more efficient than uninformed search algorithms.

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

The choice of the next node is based on ________.

A

The choice of the next node is based on THE EVALUATION FUNCTION f(n).

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

A heuristic h(n) is admissible iff …

A

UNDERESTIMATE: h(A) <= c(A; goal)

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

A heuristic h(n) is consistent iff …

A

TRIANGLE INEQUALITY: h(A) <= c(A;B) + h(B)

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

The performance of the heuristic depends on the ________.

A

… on the QUALITY OF THE HEURISTIC.

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

What are good possibilities to obtain decent heuristics?

A

Relaxed problems and pattern databases

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Greedy Best-First Search (GBFS)

A

Evaluation function is heuristic: f(n) = h(n)

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

GBFS explores ________ which is ________.

A

GBFS explores THE NODE WITH THE LOWEST HEURISTIC FUNCTION which is NOT ALWAYS OPTIMAL.

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

GBFS: Completeness

A

GBFS is complete if graph search is used

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

GBFS: Optimality

A

GBFS is NOT optimal

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

GBFS: Time/Space Complexity

A

GBFS has EXPONENTIAL time complexity O(b^m) and EXPONENTIAL space complexity O(b^m)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

A* Search (A*)

A

Evaluation function is sum of heuristic and path cost: f(n) = h(n) + c

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

A*: Completeness

A

A* is complete if costs > 0

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

A*: Optimality

A

A* is optimal iff costs > 0 && heuristic admissible

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

A*: Time/Space Complexity

A

A* has EXPONENTIAL time complexity O(b^(eps*d)) and EXPONENTIAL space complexity O(b^(eps*d))

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

The epsilon parameter for A* determines ________.

A

Epsilon determines the PERFORMANCE OF THE HEURISTIC (0 -> good / 1 -> bad).

17
Q

Main disadvantage of A* and solution

A

Possibly huge space consumption –> Use Iterative Deepening A*

18
Q

Informed search methods require a ________ that estimates ________.

A

Informed search methods require a HEURISTIC FUNCTION that estimates THE COST h(n) FROM A NODE TO THE GOAL.