search problems russell and norvig: chapter 3, sections 3.1 – 3.3
Post on 21-Dec-2015
219 views
TRANSCRIPT
Search ProblemsSearch Problems
Russell and Norvig: Chapter 3, Sections 3.1 – 3.3
Search Problems 2
Problem-Solving Problem-Solving AgentAgent
environmentagent
?
sensors
actuators
Search Problems 3
Problem-Solving AgentProblem-Solving Agent
environmentagent
?
sensors
actuators• Actions • Initial state• Goal test
Search Problems 4
문제를 어떻게 탐색 문제로
단순한 탐색 길 찾기 추적하기
규칙 문제를 탐색으로 If 비교조건 then 행동 또는 결론 우선순위 없는 수많은 규칙에서 적합한 것 찾기
고차원적 문제 Meta-knowledge 에서 찾기
Search Problems 5
State Space and Successor State Space and Successor FunctionFunction
• Actions • Initial state• Goal test
state space
successor function
Search Problems 6
Initial StateInitial State
• Actions • Initial state• Goal test
state space
successor function
Search Problems 7
Goal TestGoal Test
• Actions • Initial state• Goal test
state space
successor function
Search Problems 8
Example: 8-puzzleExample: 8-puzzle
1
2
3 4
5 6
7
8 1 2 3
4 5 6
7 8
Initial state Goal state
Search Problems 9
Example: 8-puzzleExample: 8-puzzle
1
2
3 4
5 6
7
8
1
2
3 4
5 6
7
8
1
2
3 4
5 6
78
1
2
3 4
5 6
7
8
Search Problems 10
Example: 8-puzzleExample: 8-puzzle
Size of the state space = 9!/2 = 181,440
15-puzzle .65 x 1012
24-puzzle .5 x 1025
10 millions states/sec
0.18 sec
6 days
12 billion years
Search Problems 11
Search ProblemSearch Problem
State space Initial state Successor function Goal test Path cost
Search Problems 12
Search ProblemSearch Problem
State space each state is an abstract
representation of the environment the state space is discrete
Initial state Successor function Goal test Path cost
Search Problems 13
Search ProblemSearch Problem
State space Initial state: usually the current state sometimes one or several
hypothetical states (“what if …”)
Successor function Goal test Path cost
Search Problems 14
Search ProblemSearch Problem
State space Initial state Successor function: [state subset of states] an abstract representation of the
possible actions
Goal test Path cost
Search Problems 15
Search ProblemSearch Problem
State space Initial state Successor function Goal test: usually a condition sometimes the description of a state
Path cost
Search Problems 16
Search ProblemSearch Problem
State space Initial state Successor function Goal test Path cost: [path positive number] usually, path cost = sum of step costs e.g., number of moves of the empty tile
Search Problems 17
Search of State SpaceSearch of State Space
Search Problems 18
Search of State SpaceSearch of State Space
Search Problems 19
Search State SpaceSearch State Space
Search Problems 20
Search of State SpaceSearch of State Space
Search Problems 21
Search of State SpaceSearch of State Space
Search Problems 22
Search of State SpaceSearch of State Space
search tree
Search Problems 23
Simple Agent AlgorithmSimple Agent Algorithm
Problem-Solving-Agent1. initial-state sense/read state2. goal select/read goal3. successor select/read action models4. problem (initial-state, goal,
successor)5. solution search(problem)6. perform(solution)
Search Problems 24
Example: 8-queensExample: 8-queens
Place 8 queens in a chessboard so that no two queens are in the same row, column, or diagonal.
A solution Not a solution
Search Problems 25
Example: 8-queensExample: 8-queens
Formulation #1:• States: any arrangement of 0 to 8 queens on the board• Initial state: 0 queens on the board• Successor function: add a queen in any square• Goal test: 8 queens on the board, none attacked
648 states with 8 queens
Search Problems 26
Example: 8-queensExample: 8-queensFormulation #2:• States: any arrangement of k = 0 to 8 queens in the k leftmost columns with none attacked• Initial state: 0 queens on the board• Successor function: add a queen to any square in the leftmost empty column such that it is not attacked by any other queen• Goal test: 8 queens on the board
2,067 states
Search Problems 27
실제 n-queen 문제
Neural, Genetic 또는 Heuristic 방법으로 잘 해결 최악의 경우에는 처리 불가능 실제 n 이 커지면 답이 매우 많으므로 간단한
Heuristics 로도 답을 쉽게 찾음 따라서 n 이 커도 답을 잘 찾는다고 해서 인공지능
접근방법이 문제를 해결한다는 증거는 아님 그러나 많은 실제 문제는 알고리즘에서 이야기하는
최악의 경우로는 잘 가지 않음 더구나 대부분 우리가 원하는 답은 최적이 아니라 실제
활용해서 도움이 되는 , feasible solution 을 원하므로 인공지능 기법이 효과적으로 이용될 수 있음
Search Problems 28
Example: Robot Example: Robot navigationnavigation
What is the state space?
Search Problems 29
Example: Robot Example: Robot navigationnavigation
Cost of one horizontal/vertical step = 1Cost of one diagonal step = 2
Search Problems 30
Example: Robot Example: Robot navigationnavigation
Search Problems 31
Example: Robot Example: Robot navigationnavigation
Search Problems 32
Example: Robot Example: Robot navigationnavigation
Cost of one step = ???
Search Problems 33
Example: Robot Example: Robot navigationnavigation
Search Problems 34
Example: Robot Example: Robot navigationnavigation
Search Problems 35
Example: Robot Example: Robot navigationnavigation
Cost of one step: length of segment
Search Problems 36
Example: Robot Example: Robot navigationnavigation
Search Problems 37
Example: Assembly Example: Assembly PlanningPlanning
Initial state
Goal state
Successor function:• Merge two subassemblies
Complex function:it must find if a collision-freemerging motion exists
Search Problems 38
Example: Assembly Example: Assembly PlanningPlanning
Search Problems 39
Example: Assembly Example: Assembly PlanningPlanning
Search Problems 40
Assumptions in Basic Assumptions in Basic SearchSearch
The environment is staticThe environment is discretizableThe environment is observableThe actions are deterministic
open-loop solution
Search Problems 41
Search Problem Search Problem FormulationFormulation
Real-world environment Abstraction
Search Problems 42
Search Problem Search Problem FormulationFormulation
Real-world environment Abstraction Validity:
Can the solution be executed?
Search Problems 43
Search Problem Search Problem FormulationFormulation
Real-world environment Abstraction Validity:
Can the solution be executed? Does the state space contain the
solution?
Search Problems 44
Search Problems 45
Search Problems 46
Search Problems 47
Search Problems 48
Search Problem Search Problem FormulationFormulation
Real-world environment Abstraction Validity:
Can the solution be executed? Does the state space contain the
solution? Usefulness
Is the abstract problem easier than the real-world problem?
Search Problems 49
Search Problem Search Problem FormulationFormulation
Real-world environment Abstraction Validity:
Can the solution be executed? Does the state space contain the solution?
Usefulness Is the abstract problem easier than the real-
world problem?
Without abstraction an agent would be swamped by the real world
Search Problems 50
Search Problem VariantsSearch Problem Variants
One or several initial states One or several goal states The solution is the path or a goal node In the 8-puzzle problem, it is the path
to a goal node In the 8-queen problem, it is a goal
node
Search Problems 51
Problem VariantsProblem Variants
One or several initial states One or several goal states The solution is the path or a goal node Any, or the best, or all solutions
Search Problems 52
Important ParametersImportant Parameters
Number of states in state space
8-puzzle 181,44015-puzzle .65 x 1012 24-puzzle .5 x 1025
8-queens 2,057100-queens 1052
There exist techniques to solveN-queens problems efficiently!
Stating a problem as a search problem is not always a good idea!
Search Problems 53
Important ParametersImportant Parameters
Number of states in state space Size of memory needed to store a state
Search Problems 54
Important ParametersImportant Parameters
Number of states in state space Size of memory needed to store a state Running time of the successor function
Search Problems 55
ApplicationsApplications
Route finding: airline travel, telephone/computer networks Pipe routing, VLSI routing Pharmaceutical drug design Robot motion planning Video games
Search Problems 56
Task Environment Observable Deterministic Episodic Static Discrete Agents
Crossword puzzle Fully Deterministic Sequential Static Discrete Single
Chess with a clock Fully Strategic Sequential Semi Discrete Multi
Poker Partially Strategic Sequential Static Discrete Multi
Backgammon Fully Stochastic Sequential Static Discrete Multi
Taxi driving Partially Stochastic Sequential Dynamic Continuous Multi
Medical diagnosis Partially Stochastic Sequential Dynamic Continuous Single
Image-analysis Fully Deterministic Episodic Semi Continuous Single
Part-picking robot Partially Stochastic Episodic Dynamic Continuous Single
Refinery controller Partially Stochastic Sequential Dynamic Continuous Single
Interactive English tutor Partially Stochastic Sequential Dynamic Discrete Multi
Figure 2.6 Examples of task environments and their characteristics.
Search Problems 57
SummarySummary
Problem-solving agent State space, successor function, search Examples: 8-puzzle, 8-queens, route finding, robot navigation, assembly planning Assumptions of basic search Important parameters
Search Problems 58
Future ClassesFuture Classes
Search strategies Blind strategies Heuristic strategies
Extensions Uncertainty in state sensing Uncertainty action model On-line problem solving