AI Exam for Software Dev & Computer Networking Students, CIT, Autumn 2007, Exams of Artificial Intelligence

Download AI Exam for Software Dev & Computer Networking Students, CIT, Autumn 2007 and more Exams Artificial Intelligence in PDF only on Docsity! Cork Institute of Technology 1 Bachelor of Science (Honours) in Software Development and Computer Networking – Award (NFQ Level 8) Autumn 2007 Artificial Intelligence (Time: 2 Hours) Answer any TWO questions (All questions carry equal marks). Examiners: Dr. J. Buckley Mr. A. Kinsella Mr. P. Rothwell 1. (a) Outline the physical symbol system hypothesis. [8 marks] Briefly discuss its importance. [6 marks] Briefly define a knowledge representation language. [4 marks] (b) The Frame Problem is seen as a serious impasse for logic-based knowledge representation. Fully explain the frame problem using examples. [18 marks] (c) Hubert Dreyfus criticises attempts to represent common-sense knowledge using a formal propositional knowledge representation language. Describe and explain the argument he gives in support of his criticism. [14 marks] [Total: 50 marks] 2 2. (a) List and explain the essentials of the state-space representation of search problems used in artificial intelligence. [12 marks] Define the full state-space representation that you would use in attempting to solve the following “missionaries and cannibals problem.” (Do not bother with solving the problem) [10 marks] 3 missionaries and 3 cannibals are on one side of a river. They have one boat that can carry a maximum of two people either way. The boat must be manned. Find a way to get them all to the other side without ever allowing the cannibals to outnumber the missionaries on a bank. (b) With the aid of pseudocode, fully explain the A* search algorithm. [18 marks] (c) Briefly discuss each of the following with reference to the A* algorithm: efficiency, optimality, the source of heuristics, the role of heuristics, admissible heuristics. [10 marks] [Total: 50 marks] 3. (a) Write out and fully explain Nilsson’s algorithm for searching for optimal solution trees in AND/OR graphs. [20 marks] (b) Show and explain a trace of the operation of Nilsson’s algorithm on the following AND/OR graph. The costs of the paths are shown below; assume all initial heuristic values h(n)=1. Make sure to show how the algorithm you described in part (a) achieves the results you show in part (b). [30 marks] [Total: 50 marks]