1. Pengantar Kecerdasan Buatan
Penyelesaian Masalah dengan Pencarian

2. Problem-solving Agent
Problem-solving agent decide what to do by finding sequences of action that lead to a desirable state
Problem-solving agent is a kind of goal-based agent
Oscar Karnalim, S.T., M.T.

3. How do We Solve Problem ? Define the problem Look for solution
Solve the problem
Evaluate the result
Oscar Karnalim, S.T., M.T.

4. How does Agent Work ? Goal Formulation Problem Formulation Search
Know what to do / where to go .
Problem Formulation
The process of deciding what actions and states to consider and follow goal formulation
Search
Finding solution
Solution
The result of Search Process
Execution
Remember: “Formulate, Search, Execute”
Oscar Karnalim, S.T., M.T.

5. Problem Example The Road to Bucarest
An agent enjoying his holiday in the city of Arad, Romania.
The agent’s home flight leaves tomorrow from Bucarest (the flight ticket is nonrefundable)
Oscar Karnalim, S.T., M.T.

6. Romania
Oscar Karnalim, S.T., M.T.

7. The Road to Bucharest Formulate goal: Formulate problem:
be in Bucharest
Formulate problem:
states: various cities
actions: drive between cities
Find solution:
sequence of cities,
e.g., Arad, Sibiu, Fagaras, Bucharest
Oscar Karnalim, S.T., M.T.

8. Simplified Road Map to Bucarest
Oscar Karnalim, S.T., M.T.

9. Problem Definition Initial state e.g., In(Arad)
Possible actions available or successor function
S(x) = set of action–state pairs
e.g., S(Arad) = {<Arad  Zerind, Zerind>, … }
Goal test, can be
explicit, e.g., x = "at Bucharest"
implicit, e.g., Checkmate(x)
Path cost (additive)
e.g., sum of distances, number of actions executed, etc.
c(x,a,y) is the step cost, assumed to be ≥ 0
A solution is a sequence of actions leading from the initial state to a goal state
Oscar Karnalim, S.T., M.T.

10. Problem Definition (Cont)
A Problem can be defined by four component
Initial State, e.g. In (Arad)
Action or Successor function, a description of the possible actions available to the agent
<action, successor>
e.g. {<Go(Sibiu), In(Sibiu)>, <Go(Timisoara), In(Timisoara)>,…}
Goal Test, e.g. In (Bucharest)
Path Cost, a function that assigns a numeric cost to each path
Oscar Karnalim, S.T., M.T.

11. Selecting a State Space
Real world is absurdly complex
 state space must be abstracted for problem solving
(Abstract) state = set of real states
(Abstract) action = complex combination of real actions
e.g., "Arad  Zerind" represents a complex set of possible routes, detours, rest stops, etc.
(Abstract) solution = set of real paths that are solutions in the real world
Each abstract action should be "easier" than the original problem
Oscar Karnalim, S.T., M.T.

12. The Vacuum World
Goal:
Clean up all dirt Problem Formulation: 8 possible states as shown 3 possible actions {Left, Right, Suck} Solution: State {7,8}
Oscar Karnalim, S.T., M.T.

13. Vacuum World State Space
States? dirt and robot location
Initial state? Any state
Actions? Left, Right, Suck
Goal test? no dirt at all locations
Path cost? 1 per action
Oscar Karnalim, S.T., M.T.

14. The 8-Puzzle State Space
States? locations of tiles
Initial state? Any state
Actions? move blank left, right, up, down
Goal test? = goal state (given)
Path cost? 1 per move
Oscar Karnalim, S.T., M.T.

15. Robotic Assembly State Space
States? real-valued coordinates of robot joint angles parts of the object to be assembled
Actions? continuous motions of robot joints
Goal test? complete assembly
Path cost? time to execute
Oscar Karnalim, S.T., M.T.

16. Real World Problems Route-finding problem Touring problem
Traveling salesperson problem
Robot Navigation
Automatic assembly sequencing
Protein Design
Internet Searching
Oscar Karnalim, S.T., M.T.

17. Searching for A Solution
Finding a solution  searching through state space.
Generating action sequence
Start with initial state
Check for goal state
Expand state
The process of choosing which state to expand first is called search strategy
Oscar Karnalim, S.T., M.T.

18. The Search Tree
It is helpful to think of the search process as building up a search tree.
We explore problem space with a search tree
Tree Nodes represent states
Edges represent operators
The root of the tree is the initial state
The node is expanded to generate its children
Oscar Karnalim, S.T., M.T.

19. Implementation: States vs. Nodes
A state is a (representation of) a physical configuration
A node is a data structure constituting part of a search tree includes state, parent node, action, path cost g(x), depth
Oscar Karnalim, S.T., M.T.

20. Konversi Masalah ke Pohon
Oscar Karnalim, S.T., M.T.

21. Contoh Pohon Pencarian
Oscar Karnalim, S.T., M.T.

22. Contoh Pohon Pencarian (Cont)
Oscar Karnalim, S.T., M.T.

23. Contoh Pohon Pencarian (Cont)
Oscar Karnalim, S.T., M.T.

24. Strategi Pencarian
A search strategy is defined by picking the order of node expansion
Strategies are evaluated along the following dimensions:
completeness: does it always find a solution if one exists?
time complexity: number of nodes generated
space complexity: maximum number of nodes in memory
optimality: does it always find a least-cost solution?
Time and space complexity are measured in terms of
b: maximum branching factor of the search tree
d: depth of the least-cost solution
m: maximum depth of the state space (may be ∞)
Oscar Karnalim, S.T., M.T.

25. Uninformed Search
It have no information about the number of step or the path cost from the current state to the goal, can only distinguish between goal and non-goal state.
For the same reason it is also called blind search
Oscar Karnalim, S.T., M.T.

26. Uninformed Search Strategies
Breadth-first search
Uniform-cost search
Depth-first search
Depth-limited search
Iterative deepening search
Oscar Karnalim, S.T., M.T.

27. Breadth-first Search Expand shallowest unexpanded node
fringe is a FIFO queue, i.e., new successors go at end
Oscar Karnalim, S.T., M.T.

28. Breadth-first Search (Cont)
Oscar Karnalim, S.T., M.T.

29. Properties of Breadth-first Search
Complete? Yes (if b is finite)
Time? 1+b+b2+b3+… +bd + b(bd-1) = O(bd+1)
Space? O(bd+1) (keeps every node in memory)
Optimal? Yes (if cost = 1 per step)
Space is the bigger problem (more than time)
Oscar Karnalim, S.T., M.T.

30. Uniform-cost Search Expand least-cost unexpanded node Implementation:
fringe = queue ordered by path cost
Equivalent to breadth-first if step costs all equal
Complete? Yes, if step cost ≥ ε
Time? # of nodes with g ≤ cost of optimal solution, O(bceiling(C*/ ε)) where C* is the cost of the optimal solution
Space? # of nodes with g ≤ cost of optimal solution, O(bceiling(C*/ ε))
Optimal? Yes – nodes expanded in increasing order of g(n)
Oscar Karnalim, S.T., M.T.

31. Uniform-cost Search (Cont)
Oscar Karnalim, S.T., M.T.

32. Depth-first Search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front
Oscar Karnalim, S.T., M.T.

33. Depth-first Search (Cont)
Oscar Karnalim, S.T., M.T.

34. Depth-first Search (Cont)
Oscar Karnalim, S.T., M.T.

35. Properties of Depth-first Search
Complete? No: fails in infinite-depth spaces, spaces with loops
Modify to avoid repeated states along path
 complete in finite spaces
Time? O(bm): terrible if m is much larger than d
but if solutions are dense, may be much faster than breadth-first
Space? O(bm), i.e., linear space!
Optimal? No
Oscar Karnalim, S.T., M.T.

36. Depth-limited Search Depth-first search with depth limit
i.e., nodes at depth l have no successors
Oscar Karnalim, S.T., M.T.

37. Iterative Deepening Search
Depth-first search by trying all possible depth limits. First depth=0, then depth=1, then depth=2, and so on.
Combines the benefits of BFS and DFS
Oscar Karnalim, S.T., M.T.

38. Iterative Deepening Search with i=0
Oscar Karnalim, S.T., M.T.

39. Iterative Deepening Search with i=1
Oscar Karnalim, S.T., M.T.

40. Iterative Deepening Search with i=2
Oscar Karnalim, S.T., M.T.

41. Iterative Deepening Search with i=3
Oscar Karnalim, S.T., M.T.

42. Iterative Deepening Search
Number of nodes generated in a depth-limited search to depth d with branching factor b:
NDLS = b0 + b1 + b2 + … + bd-2 + bd-1 + bd
Number of nodes generated in an iterative deepening search to depth d with branching factor b:
NIDS = (d+1)b0 + d b^1 + (d-1)b^2 + … + 3bd-2 +2bd-1 + 1bd
For b = 10, d = 5,
NDLS = , , ,000 = 111,111
NIDS = , , ,000 = 123,456
Overhead = (123, ,111)/111,111 = 11%
Oscar Karnalim, S.T., M.T.

43. Properties of Iterative Deepening Search
Complete? Yes
Time? (d+1)b0 + d b1 + (d-1)b2 + … + bd = O(bd)
Space? O(bd)
Optimal? Yes, if step cost = 1
Oscar Karnalim, S.T., M.T.

44. Summary of Algorithms
Oscar Karnalim, S.T., M.T.

45. Bidirectional Search
Oscar Karnalim, S.T., M.T.

46. Repeated States
Failure to detect repeated states can turn a linear problem into an exponential one!
Oscar Karnalim, S.T., M.T.

47. Avoiding Repeated States
Three ways to deal with repeated states, in increasing order of effectiveness and computational overhead: Do not return to the state you just came from Do not create paths with cycles in the space state Do not generate any state that was ever generated before [this require every generated state to be kept in memory, resulting in space complexity O(bd)]
Oscar Karnalim, S.T., M.T.

48. Referensi
Russell, Stuart J. & Norvig, Peter. Artificial Intelligence: A Modern Approach (3rd ed.). Prentice Hall. 2009
Luger, George F. Artificial Intelligence: Structures and Strategies for Complex Problem Solving. 6th Edition. Addison Wesley
Watson, Mark., Practical Artificial Intelligence Programming in Java, Open Content – Free eBook (CC License), 2005.
Oscar Karnalim, S.T., M.T.

49. Kuis2 Apakah agent itu? Apakah yang dimasudkan dengan PEAS?
Berikan contoh sebuah agent lengkap dengan PEAS-nya.
Berikan penjelasan singkat tentang 4 macam jenis agent.
Oscar Karnalim, S.T., M.T.