BITM Artificial Intelligence

Do you support biasness in AI? Give reason.

How do you measure utility in agent?

What is adversarial search?

List any two differences between simple reflexive and model based agent.

What is a production system in the context of rule-based knowledge representation?

List the two types of quantifiers.

Define tautology.

What is reinforcement learning?

When do your search is complete?

What is well defined problem?

If AI passes the Turing test but has zero consciousness, did it actually pass the test? Justify your option.

Write PEAS description for virtual shopping assistant.

List and discuss the operators of genetic algorithm.

Define fuzzy sets? How do you define membership in fuzzy set?

Discuss the pragmatic ambiguity in NLP. How can we solve it?

Explain components of machine vision system.

How do you represent knowledge using semantic nets and scripts? Explain with example.

How do you train Hebbian learning to simulate the logical OR function? Illustrate.

What is expert system? Explain the expert system with block diagram.

Explain the concept of Back-propagation learning in neural networks. How does it improve the accuracy of the model during training?

Why do we need depth limited search and what are its limitations? Find the global maximum among two scenarios, one starting at \(x = 5\) and other starting at \(x = 3\) using Hill climbing search for the followings given information.

From any position x, you can only move to your immediate left neighbour \((x - 1)\) or immediate right neighbor \((x + 1)\).

Define proposition, unification and lifting. An engineer is analyzing a smart power grid under three binary random variables. Now given the following full joint distribution table, answer the given questions.

a. What is the total prior probability that the grid experience is a solar urge?

b. If there is no solar surge, what is the probability that the system is facing a failure risk?