170 likes | 316 Views
Artificial Intelligence Lecture. Md. Morshedul Islam Assistant Professor Department of Computer Science & Engineering Bangladesh University of Business and Technology (BUBT). Contents. Informed Search Strategies Heuristic Information Hill Climbing Methods Best-First Search
E N D
Artificial IntelligenceLecture Md. Morshedul Islam Assistant Professor Department of Computer Science & EngineeringBangladesh University of Business and Technology (BUBT) Artificial Intelligence, Lecturer #14
Contents • Informed Search Strategies • Heuristic Information • Hill Climbing Methods • Best-First Search • Optimal Search and A* • Branch-and-Bound Search • Iterative Deepening A* • Memory-bounded Heuristic Search Artificial Intelligence, Lecturer #14
S 21 28 24 18 19 23 23 16 16 21 25 19 9 25 25 11 22 27 25 25 20 25 Hill Climbing Methods:An Example Artificial Intelligence, Lecturer #14
S 20 22 28 18 12 22 20 25 15 18 16 5 12 8 10 12 10 10 15 5 2 8 Best-First Search: An Example L1: S20, S22, S28 L2: S35, S36, S38 L3: S40, S45 g Artificial Intelligence, Lecturer #14
Optimal Search and A* • The A* algorithm is a specialization of best-first-search • It provides general guidelines with which to estimate goal distances for general search graphs. • At each node along a path to the goal, the A* algorithm generates all successors nodes and computes the distance (cost) from the start node to a goal node through each of the successor. • It then chooses the successor with the shortest distance for expansion. • The successor for this node are then generated. • Node are labeled with f(n) = g(n)+h(n) Artificial Intelligence, Lecturer #14
O 71 N 151 87 I 75 Z 140 F 92 V A S 99 118 142 80 R 211 T 97 U P 111 98 L 101 H 70 146 M 85 86 B 75 138 120 90 E D C G Optimal Search and A*: ExampleRoad Map of a City Artificial Intelligence, Lecturer #14
Optimal Search and A*: Example Values of hSLD-straight-line distance to B Artificial Intelligence, Lecturer #14
A (a) Initial State: 366+0 A (b) Expanding A: S T Z 393=140+253 447=118+329 449=75+374 Optimal Search and A*: ExampleNode are labeled with f(n) = g(n)+h(n) Artificial Intelligence, Lecturer #14
A (c) Expanding S: 140 S T Z 140 99 447=118+329 449=75+374 151 80 A F O R 415=239+176, 413=220+193 646=280+366, 671=291+380 Optimal Search and A*: ExampleNode are labeled with f(n) = g(n)+h(n) Artificial Intelligence, Lecturer #14
A (d) Expanding R: 140 S T Z 140 99 447=118+329 449=75+374 151 80 A F O R 415=239+176, 80 646=280+366, 671=291+380 97 146 S P C 553=300+253 417=317+100 526=366+160 Optimal Search and A*: ExampleNode are labeled with f(n) = g(n)+h(n) Artificial Intelligence, Lecturer #14
A (e) Expanding F: 140 S T Z 140 99 447=118+329 449=75+374 151 80 A F O R 80 646=280+366, 671=291+380 97 211 99 146 S S B P C 591=338+253 553=300+253 450=450+0 417=317+100 526=366+160 Optimal Search and A*: ExampleNode are labeled with f(n) = g(n)+h(n) Artificial Intelligence, Lecturer #14
A (f) Expanding P: 140 S T Z 140 99 447=118+329 449=75+374 151 80 A F O R 80 646=280+366, 671=291+380 97 211 99 146 S S B P C 591=338+253 553=300+253 450=450+0 526=366+160 138 97 101 B C R 418=418+0 615=455+160 607=414+193 Optimal Search and A*: Example Artificial Intelligence, Lecturer #14
A 366 329 374 253 S T Z A F O R 366 176 380 193 S B 253 0 Greedy Best-first Search: ExampleNode are labeled with f(n) =h(n) Artificial Intelligence, Lecturer #14
Branch-and-Bound Search • This strategy saves all paths lengths (or costs) from a node to all generated nodes and chooses the shortest path for farther expansion. • It then compares the new path length with all old ones and again chose the shortest path for expansion. • In this method, a lowest cost path will be found. • Expensive in case of computing and remembering all computing paths Artificial Intelligence, Lecturer #14
S 12 14 15 9 11 11 13 8 11 12 14 12 9 10 1 9 1 13 15 5 10 9 G Branch-and-Bound Search: Example L1: S12,S14, S15 L2: S12+11, S12+14, S12+12; S14+11, S14+9; S15+13, S15+8, S15+11 L3: S23+12, S23+13, S26+13, S26+9, S24+9, S24+10; S25+10, S25+9, S23+9; S28+9, S28+15, S23+1, S26+1, S26+5
Recommended Textbooks • [Negnevitsky, 2001] M. Negnevitsky “ Artificial Intelligence: A guide to Intelligent Systems”, Pearson Education Limited, England, 2002. • [Russel, 2003] S. Russell and P. Norvig Artificial Intelligence: A Modern Approach Prentice Hall, 2003, Second Edition • [Patterson, 1990] D. W. Patterson, “Introduction to Artificial Intelligence and Expert Systems”, Prentice-Hall Inc., Englewood Cliffs, N.J, USA, 1990. • [Minsky, 1974] M. Minsky “A Framework for Representing Knowledge”, MIT-AI Laboratory Memo 306, 1974. Artificial Intelligence, Lecturer #14
End of Presentation Questions or Suggestions? Thanks to all !!! Artificial Intelligence, Lecturer #14