EC-802-A AI and Signal Processing - May 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. 7 Marks
   aWhat do you understand by Artificial Intelligence? How Artificial intelligence, Machine Learning, and Deep Learning differ from each other?
   How to answer

   • Define Artificial Intelligence
   • Define Machine Learning as subset of AI and Deep Learning as subset of ML
   • Differentiate on basis of scope, data requirement, computation and examples
   • Give examples for each
2. 7 Marks
   bWhat are the different types of Machine Learning? How is Machine Learning related to Artificial Intelligence?
   How to answer

   • List types of Machine Learning: supervised, unsupervised, reinforcement
   • Brief characteristics of each type
   • Explain ML as subset/approach to realize AI
   • Give examples of relation
3. 7 Marks
   aWhat are the Examples of AI in real life? Also explain Differentiate between Natural (Human) Intelligence and Artificial Intelligence.
   How to answer

   • List real-life examples of AI: virtual assistants, recommendation, autonomous vehicles etc.
   • Define Natural (Human) Intelligence
   • Differentiate on basis of origin, learning, speed, accuracy
   • Conclude with comparison table
4. 7 Marks
   bHow is AI related to Knowledge? Differentiate between Declarative and Procedural knowledge.
   How to answer

   • Explain role of knowledge in AI systems
   • Define declarative knowledge with examples
   • Define procedural knowledge with examples
   • Differentiate in table: representation, use, examples
5. 7 Marks
   aWhat is Utility theory and its importance in AI? Explain with the help of suitable examples.
   How to answer

   • Define utility theory and utility function
   • Explain importance in rational agent decision making under uncertainty
   • Give suitable examples e.g., game, robot navigation
   • Mention expected utility maximization
6. 7 Marks
   bExplain Backward and forward Chaining, with example in logic representation. Also mention advantages and disadvantages of both the algorithms.
   How to answer

   • Define forward chaining with logic example
   • Define backward chaining with logic example
   • Compare working, direction and use cases
   • List advantages and disadvantages of both
7. 7 Marks
   aExplain in detail Markov Model and its applications in Artificial Intelligence.
   How to answer

   • Define Markov model and Markov property
   • Explain types: Markov chain, HMM
   • Give transition probability formulation
   • List AI applications
8. 7 Marks
   bWhat is Bayesian Network or Belief Network? Explain its importance with the help of an example.
   How to answer

   • Define Bayesian / Belief Network as DAG with conditional probabilities
   • Explain nodes, edges, CPT with example
   • Show inference / importance
   • Give example e.g., medical diagnosis
9. 7 Marks
   aExplain Machine Learning. Illustrate learning model. Mention some factors that affect the learning.
   How to answer

   • Define Machine Learning
   • Illustrate learning model with block diagram description
   • Explain components: environment, learner, knowledge base
   • List factors affecting learning
10. 7 Marks
    bCompare the Decision tree method with Naive Baye's Learning.
    How to answer

    • Define Decision tree method
    • Define Naive Bayes learning
    • Compare on principle, handling data, assumptions
    • Conclude suitability
11. 7 Marks
    aCompute the eight-point DFT circular convolution for the following sequences. $x_{2}(n) = \sin 3\pi n/8$
    How to answer

    • State DFT and circular convolution definitions
    • Use DFT property to compute convolution via IDFT of product
    • Evaluate $x_2(n)=\sin 3\pi n/8$ for n=0..7
    • Obtain 8-point result
12. 7 Marks
    bState and prove the:

    1. i)Circular convolution property of DFT
    2. ii)DFT of Real and even sequence
    How to answer

    • State circular convolution property: DFT of circular convolution = product of DFTs
    • Prove using DFT summation and periodicity
    • State DFT of real and even sequence property
    • Prove symmetry: X(k) real and even
13. 7 Marks
    aWhat do you mean by "Quadrature mirror filter banks"? Also explain How many types of filter banks are there in multi rate signal processing?
    How to answer

    • Define Quadrature Mirror Filter banks and alias cancellation
    • Explain two-channel QMF structure
    • List types of filter banks in multirate processing
    • Mention applications
14. 7 Marks
    bThe DFT of a 5-point signal $x(n), 0 \le n \le 4$ is $X(k) = [5, 6, 1, 2, 9], 0 \le k \le 4$. A new signal $g(n)$ is defined by $g(n) := W_{5}^{-2n} x(n), 0 \le n \le 4$. What are the DFT coefficients $G(k)$ of the signal $g(n)$, for $0 \le k \le 4$?
    How to answer

    • Recall DFT $X(k)$ and twiddle factor $W_N$
    • Apply modulation / frequency-shift property for $g(n)=W_5^{-2n}x(n)$
    • Map $G(k)=X((k+2))_5$ circular shift
    • Compute values for $k=0..4$
15. 14 Marks
    Write short notes on: (any two)

    1. a)Spline interpolation
    2. b)Computationally efficient sampling rate converters
    3. c)Parallel all pass realization of IIR transfer functions
    4. d)Basic FIR/IIR filter structures
    How to answer

    • Select any two topics from a-d
    • Define each with key principles/formulas
    • Explain structures / steps with examples
    • Mention applications / advantages