AD-304 Artificial Intelligence - Jun 2023 Question Paper

Under each question, How to answer lists the points an answer should cover, outlined by CampusPrep from the syllabus. Check the details against your textbook.

1. 10 Marks
   aExplain the Best First Search algorithm in detail?
   How to answer

   • Concept of Best First Search combining BFS and DFS using an evaluation function $f(n) = h(n)$
   • Algorithm steps using OPEN and CLOSED lists
   • Priority queue implementation
   • Example graph/tree search showing node expansion
   • Time and space complexity, advantages and disadvantages
2. 4 Marks
   bList and explain the goals of Artificial Intelligence (AI).
   How to answer

   • List primary goals: create expert systems, implement human intelligence in machines
   • Reasoning, learning, problem-solving, perception, linguistic intelligence
   • Brief explanation of scientific and engineering objectives
3. 9 Marks
   aExplain how does predicate logic help in knowledge representation in AI?
   How to answer

   • Definition of Predicate Logic / First-Order Logic (FOL)
   • Components: constants, variables, predicates, functions, connectives, quantifiers ($\forall, \exists$)
   • Role in knowledge representation: expressiveness, formal semantics, inference capabilities
   • Concrete examples translating English sentences to FOL
4. 5 Marks
   bWhat is AI Inference? Explain.
   How to answer

   • Definition of inference: deriving new facts/conclusions from known premises
   • Role in knowledge-based systems and inference engines
   • Common rules of inference (e.g., Modus Ponens, Resolution)
   • Types of inference: deductive, inductive, abductive
5. 8 Marks
   aDefine Baye's Theorem? Explain the applications of Baye's Theorem.
   How to answer

   • Mathematical formulation: $P(A|B) = \frac{P(B|A)P(A)}{P(B)}$
   • Explanation of terms: prior $P(A)$, likelihood $P(B|A)$, evidence $P(B)$, posterior $P(A|B)$
   • Applications in AI: spam filtering, medical diagnosis, Naive Bayes classifier, uncertainty handling
6. 6 Marks
   bExplain the causes of uncertainty in AI?
   How to answer

   • Definition of uncertainty in intelligent agents
   • Causes: incomplete information, noisy data/sensors, contradictory data, vague/fuzzy statements, dynamic environments
   • Brief mention of handling methods: probabilistic reasoning, fuzzy logic
7. 7 Marks
   aGive an overview of applications of Natural Language Processing?
   How to answer

   • Introduction to Natural Language Processing (NLP)
   • Key applications: machine translation (e.g. Google Translate), sentiment analysis, chatbots/conversational agents
   • Information extraction, speech recognition, spell-checking and text summarization
8. 7 Marks
   bDiscuss Mini-Max algorithm steps in detail?
   How to answer

   • Concept of 2-player zero-sum games (MAX and MIN)
   • Recursive formulation / algorithm steps from terminal state values up to root
   • Working through MAX layers (maximizing utility) and MIN layers (minimizing utility)
   • Illustrative game tree example with values propagated
   • Complexity analysis
9. 7 Marks
   aList and explain the components and capabilities of Expert System?
   How to answer

   • Block diagram of an Expert System
   • Components: Knowledge Base, Inference Engine, User Interface, Explanation Facility, Knowledge Acquisition Subsystem
   • Capabilities: advising, demonstrating, explaining decisions, solving complex domain-specific problems
10. 7 Marks
    bDiscuss the usage of backward chaining in Inference Engine?
    How to answer

    • Definition of backward chaining as goal-driven reasoning
    • Process: starting from hypothesis/goal, checking rules with that conclusion, recursively proving premises
    • Usage in inference engines for diagnostic and debugging systems (e.g. MYCIN)
    • Advantages in query-driven applications
11. 8 Marks
    aDiscuss A* algorithm in detail?
    How to answer

    • Evaluation function: $f(n) = g(n) + h(n)$
    • Algorithm steps using OPEN and CLOSED lists
    • Conditions for optimality: admissibility ($h(n) \le h^*(n)$) and consistency
    • Comparison with Greedy Best First Search and Dijkstra's algorithm
    • Time and space complexity
12. 6 Marks
    bExplain the various problems in knowledge representation?
    How to answer

    • Key requirements/issues: representational adequacy, inferential adequacy, inferential efficiency, acquisitional efficiency
    • Classic problems: Frame problem, Qualification problem, Ramification problem
    • Handling granularity, uncertainty, and non-monotonicity
13. 7 Marks
    aWrite a short note on backward reasoning?
    How to answer

    • Definition of backward reasoning (backward chaining)
    • Starting with a goal state and working backward to initial state/premises
    • Use in automated theorem proving, Prolog, diagnostic systems
    • Comparison with forward reasoning
14. 7 Marks
    bWhat is pruning? Discuss alpha-beta pruning method?
    How to answer

    • Definition of pruning in search trees: eliminating branches that cannot influence the final decision
    • Parameters: $\alpha$ (best value for MAX) and $\beta$ (best value for MIN)
    • Pruning condition: $\alpha \ge \beta$
    • Alpha cutoff vs beta cutoff explanation
    • Example tree diagram with pruned subtrees and complexity improvement to $O(b^{d/2})$
15. 14 Marks
    Explain any two of the following:

    1. a)AI Applications
    2. b)Refutation in AI
    3. c)Conceptual Dependency
    4. d)Benefits of Expert System
    How to answer

    • For (a) AI Applications: Healthcare, finance, autonomous driving, robotics, e-commerce
    • For (b) Refutation in AI: Proof by contradiction using resolution, converting clauses to CNF, showing unsatisfiability with empty clause
    • For (c) Conceptual Dependency (CD): Schank's theory, primitive ACTs (TRANS, PTRANS, etc.), representation of sentence meanings independent of words used
    • For (d) Benefits of Expert System: High availability, low cost of expertise, permanence, fast response, unfailing reliability without bias