Informed Search Flashcards

1
Q

What is a heuristic?

A

A rule or principle used to guide a search algorithm

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

What does a heuristic search provide?

A

A workable solution in a reasonable time.

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

Where do heuristics usually come from?

A

Human experience

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

What are the basic requirements for heuristic evaluation

A
  • Computation must not take long
  • Accurate
  • Close to perfect
  • Optimistic (think it is closer to a solution than it actually is)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

How does a hill climbing algorithm consider the states?

A

As if they are all laid on the surface of a landscape

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

How does a hill climbing algorithm work?

A

Starts with current state = initial state
Until current state = goal state or no change to current state:
- Get successors of current state and apply eval. func. to all
- If one of the successors has a better score, set current state to that successor.

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

Does a hill climb algorithm retain the paths?

A

No - Little memory used

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

What is a hill climbing algorithm useful for?

A

Useful for more practical problems where the state description holds all the information needed for a solution.

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

Can a hill climbing algorithm find reasonable solutions in a large or infinite state space?

A

Yes.

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

What are some issues with hill climbing algorithms?

A
  • It may find a local minimum over a global one.
  • There may be a plateau
  • There may be ridges
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What is a plateau?

A

An area of state space where evaluation function is essentially flat
Means algorithms will go on a random walk

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

What is a ridge?

A

They have steeply sloping sides
Cause problems where states along ridge are not directly connected

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

What is the general formula for a best - first search?

A

f(n) = g(n) + h(n)
where g(n) is the cost function
and h(n) is the heuristic function

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

What does a best first algorithm use to search the tree?

A

It uses the frontier, meaning it tries all paths

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

What data structure does a best first search use?

A

A priority queue to order states on basis of how promising they are

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

How does a breadth first algorithm work?

A

Starts with priority queue = initial state
while the priority queue is not empty:
- Remove best state from pq.
- If goal node, success, otherwise find its children
- Apply eval. func. to children and add them to pq.