Punjab Technical University 2005 M.Tech Computer Science ARTIFICIAL INTELLIGENCE CS 502 - Question Paper

ARTIFICIAL INTELLIGENCE CS 502 M Tech second Sem May 2k5

Max Marks 100

Note: Attempt any 5 ques.. All ques. carry equal marks.

1. What are the different heuristic techniques? discuss how they are various from the solution guaranteed techniques.

2. (a) Derive a parse tree for the sentence “RAM hated the Loin” where he subsequent rules are used:

S - > NP VP
NP - > N
NP - > DET N
VP - > V NP
DET - > the
V - > Hates
N - > Ram | loin
(b) Under what conditions would it make sense to use both forward and backward chaining in expert system. discuss every of the chaining methods. provide an example where both are used.

3. (a) discuss the subsequent things in predicate calculus:
(i) Every person has a head
(ii) A car without wheels is not valuable
(iii)If there is something to your left and something to your right then you are crowded.
(iv) To make brownies, follow the direction on the packages.

(b) How scripts are used to represent the knowledge? discuss with example.

4 (a) Use semantic net to represent the subsequent fact.

“Every dog has bitten the postman”

(b) What are the different methods of knowledge acquisition used inexpert system? discuss with examples.

5. (a) define a method in which rules could be grouped or organized in a knowledge base to decrease the amount of search needed during the matching part of the inference cycle.

(b) elaborate the characteristics of an AI problem? explain with the help of travelling salesman issue.

6. (a) discuss why you think that associative networks were never very popular form of knowledge representations in expert system architectures.

(b) How is machine learning distinguished from general knowledge acquisition?

7. What is an inference engine in an expert system? elaborate the different functions of an inference engine? discuss this with situation to MYCIN.

8. Write short notes on the following:
(a) Manufacturing applications in AI.
(b) Principle of neural Networks.

Earning:   Approval pending.