Search Flashcards

1
Q

Finding Optimal Solutions to Rubik’s Cube?

A

The algorithm
used is iterative-deepening-A* (IDA*), with a lower- bound heuristic function based on large memory-based
lookup tables, or \pattern databases”

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

Definition of a problem

A
  • Initial State
  • Actions(s): Possible actions from s to change state
  • Result(s,a): The next state from performing action a on state s
  • GoalTest(s): Returns true if reach goal state
  • Path Cost
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Which search methods are complete and optimal

A
  • Breath First Search and Cheapest cost

- Depth first is not

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

What knowledge is most useful in search?

A
  • Estimate from the start state to the goal. A good aprox is the linear distance between them
  • Greedy best-first search: expands the path that its estimate is closest.
  • The best algorithm is a combination of greedy and uniform cost
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

A* Search

A
  • Minimum value of f = g + h
  • g(path) = path cost
  • h(path) = h(s) = estimated distance to goal
  • Finds the shortest length path while expanding minimum number of paths possible.
  • g keeps path short
  • h keeps us focused on the goal
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

When A* finds the lowest cost path?

A
  • h(s) < true cost
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What is an admissible heuristic?

A
  • heuristic function is said to be admissible if it never overestimates the cost of reaching the goal
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

How to develop automatically a good heuristic?

A
  • Relaxing the problem by eliminating one of the constraints
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

When does problem solving works?

A
  • Fully observable: must see the initial state
  • Known: must know the set of available actions
  • Discrete: Must be a finite number of actions to choose
  • Deterministic: Must know the result of an action
  • Static: only our actions can change the world
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

How to implements path in a computer?

A
  • nodes: state end of path, action to get there, total cost, parent node
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

How to implement the frontier?

A
Operations:
- Remove best item
- Add new
- Membership test
Implementation:
- Priority queue
Representation:
- Set
Build:
- Hash table
- tree
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

How to implement the explored list?

A
Operations:
- Add new
- Membership test
Representation:
- Single Set
Build:
- Hash table
- tree
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Uniform Cost Search

A
  • It is cheapest first guaranties that it finds the least total cost
How well did you know this?
1
Not at all
2
3
4
5
Perfectly