1 / 113

COMP 8620 Advanced Topics in AI

COMP 8620 Advanced Topics in AI. Lecturers: Philip Kilby & Jinbo Huang. Part 1: Search Lecturer: Dr Philip Kilby Philip.Kilby@nicta.com.au Weeks 1-7 (Week 7 is assignment seminars) Part 2: Probabilistic Reasoning with Bayesian networks Lecturer: Dr Jinbo Huang

carolinep
Download Presentation

COMP 8620 Advanced Topics in AI

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. COMP 8620Advanced Topics in AI Lecturers: Philip Kilby & Jinbo Huang

  2. Part 1: Search Lecturer: Dr Philip Kilby Philip.Kilby@nicta.com.au Weeks 1-7(Week 7 is assignment seminars) Part 2: Probabilistic Reasoning with Bayesian networks Lecturer: Dr Jinbo Huang Jinbo.Huang@nicta.com.au Weeks 8-13

  3. 1-2 Introduction Systematic Search 3-4 Linear Programming / Integer Programming / Branch and Bound 5-6 Neighbourhood-based methods:  Simulated Annealing Tabu Search  Genetic Algorithms 7-8 Constraint Programming (Guest lecturer: Jason Li) 7-8 Constraint Programming (Guest lecturer: Jason Li) 9-10 Case Studies:  TSP Planning (Jussi Rintanen) 11 Dynamic Programming 12 Qualitative Reasoning (Guest lecturer: Jason Li) 13-14 Seminars (Guest Lecturers: You lot!) Course Outline – Part 1 Search – Lecture 1-2

  4. Assessment • 40% Exam • 60% Assignment • Search • 1 x Assignment (next week) 15% • 1 x Seminar 15% • Reasoning • 3 x Assignment (10% each) (tentative) Search – Lecture 1-2

  5. What is Search? • Search finds a solution to a problem • Landscape of solutions • Some are “good” • Lots (and lots and lots) are “bad” How do we find the good ones? Search – Lecture 1-2

  6. What is Search? • Landscape can continuous or discrete • (We will only deal with discrete) • Landscape can be (approximated by?) a tree • Landscape can be (approximated by?) a graph Search – Lecture 1-2

  7. What is Search? X + 2 * Y < 3, X  [0,5] Y  [1,5] Search – Lecture 1-2

  8. X=2Y=4 X=0Y=4 X=1Y=1 X=1Y=2 X=1Y=3 X=1Y=4 X=1Y=5 X=0Y=2 X=0Y=3 X=2Y=1 X=0Y=5 X=0Y=1 X=0Y=0 X=2Y=5 X=5Y=2 X=5Y=1 X=5Y=3 X=2Y=2 X=3Y=5 X=3Y=4 X=3Y=3 X=3Y=2 X=4Y=5 X=2Y=3 X=3Y=1 X=4Y=4 X=4Y=3 X=4Y=2 X=4Y=1 X=5Y=4 X=5Y=5 X=0Y=1 What is Search? X + 2 * Y < 3, X  [0,5] Y  [1,5] Search – Lecture 1-2

  9. X=0 X=1 X=2 X=3 X=4 X=5 What is Search? X + 2 * Y < 3, X  [0,5] Y  [1,5] Y=1 Y=2 Y=3 Y=4 Y=5 Search – Lecture 1-2

  10. Y=1 Y=2 Y=3 Y=4 Y=5 X=0 X=1 X=2 X=3 X=4 What is Search? X + 2 * Y < 3, X  [0,5] Y  [1,5] X=5 Search – Lecture 1-2

  11. What is Search? X + 2 * Y < 3, X  [0,5] Y  [1,5] X>0 X=0 Y>1 Y=1 Search – Lecture 1-2

  12. What is Search? • Different flavours • Find a solution (Satisfaction/Decision) • Find the best solution (Combinatorial Optimisation) • Decision problems: • Search variant: Find a solution (or show none exists) • Decision variant: Does a solution exist? Search – Lecture 1-2

  13. What is Search • Constructive Search – Find a solution by construction • Local Search – Given a solution, explore the “neighbourhood” of that solution to find other (possibly better) solutions Search – Lecture 1-2

  14. What is Search? Anytime algorithm (for optimisation prob): • After a (short) startup time, an answer is available • The longer the alg is run, the better the solution • Quality guarantee may be available (e.g. solution is within 5% of optimal) One-shot algorithm • Answer only available at completion Search – Lecture 1-2

  15. What is Search? A problem is defined by • Initial state • Successor function (an action leading from one state to another) • Goal test • Path cost • A solution is a sequence of actions leading from the start state to a goal state Search – Lecture 1-2

  16. Example: Romania Search – Lecture 1-2

  17. Problem formulation • A problem is defined by four items: • initial state • e.g., “at Arad” • successor function S(x) • set of action–state pairs • e.g., S(Arad) = {<Arad → Zerind, at Zerind>, . . .} • goal test e.g., x = “at Bucharest” • can be implicit, e.g., HasAirport(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 Search – Lecture 1-2

  18. Tree Search functionTree-Search (problem, fringe) returns a solution, or failure fringe ← Insert (Make-Node (Initial-State (problem)), fringe) loop do if Is-Empty(fringe)then return failure node ← Remove-Front (fringe) ifGoal-Test (problem, State[node]) then return node fringe ← InsertAll (Expand (node, problem), fringe) functionExpand (node, problem) returns a set of nodes successors ← the empty set foreach (action, result) inSuccessor-Fn (problem, State[node]) do s ← a new Node Parent-Node[s] ← node; Action[s] ← action; State[s] ← result Path-Cost[s] ← Path-Cost[node] + Step-Cost(State[node], action, result) Depth[s] ← Depth[node] + 1 successors ← Insert (s) return successors Search – Lecture 1-2

  19. Breadth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  20. Breadth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  21. Breadth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  22. Breadth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  23. Breadth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  24. Breadth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  25. Breadth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  26. Breadth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  27. Breadth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  28. Breadth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  29. Breadth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  30. Breadth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  31. Breadth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  32. Breadth-first Search A B C D E F G !! H I L M O Search – Lecture 1-2

  33. Depth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  34. Depth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  35. Depth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  36. Depth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  37. Depth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  38. Depth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  39. Depth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  40. Depth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  41. Depth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  42. Depth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  43. Depth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  44. Depth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  45. Depth-first Search A B C D E F G H I L M O Search – Lecture 1-2

  46. Breadth-first Search A B C D E F G !! H I L M O Search – Lecture 1-2

  47. Iterative Deepening DFS Search A B C D E F G H I L M O Search – Lecture 1-2

  48. Iterative Deepening DFS Search A B C D E F G H I L M O Search – Lecture 1-2

  49. Iterative Deepening DFS Search A B C D E F G H I L M O Search – Lecture 1-2

  50. Iterative Deepening DFS Search A B C D E F G H I L M O Search – Lecture 1-2

More Related