IT-701 Soft Computing - Nov 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
   aDefine Neural Network. Explain with a neat sketch Biological neural network.
   How to answer

   • Definition of neural network as interconnected processing elements
   • Structure of biological neuron - dendrites, soma, axon, synapses
   • Neat labelled sketch of biological neural network
   • Function of each component in signal processing
2. 7 Marks
   bDiscuss the flow of information in neural cell.
   How to answer

   • Dendrites receive incoming signals
   • Summation and processing in soma/cell body
   • Axon transmission and synaptic output
   • Flow: reception -> integration -> conduction -> transmission
3. 7 Marks
   aWrite short notes on evolution of neural networks.
   How to answer

   • Early biological inspiration - McCulloch-Pitts, Hebb
   • Perceptron, Adaline, backpropagation milestones
   • Neocognitron to deep learning evolution
   • Generations of neural networks summary
4. 7 Marks
   bDiscuss in brief perceptron training algorithm for single output class.
   How to answer

   • Perceptron model for single output classification
   • Weight initialization and learning rate
   • Training steps: compute output, update weights
   • Stopping condition and convergence
5. 7 Marks
   aExplain with a neat sketch the architecture of Adaptive Linear Neuron (Adaline).
   How to answer

   • Adaline architecture: inputs, weights, bias, linear combiner
   • Neat labelled sketch with summation and activation
   • Difference from perceptron - linear output
   • Signal flow explanation
6. 7 Marks
   bWrite and explain the training algorithm and testing algorithm for Adaline.
   How to answer

   • LMS / Delta rule for training
   • Steps of training algorithm with weight update formula
   • Testing procedure for Adaline
   • Convergence and error minimization
7. 7 Marks
   aExplain with a neat sketch the architecture of Kohonen self-organizing map and also draw a flow chart.
   How to answer

   • Kohonen SOM architecture: input layer and Kohonen layer
   • Neat sketch with lateral connections and neighbourhood
   • Training flow chart steps
   • Winner selection and weight update
8. 7 Marks
   bDiscuss in detail about Forward-Only Counter propagation Network.
   How to answer

   • Concept of counterpropagation network
   • Forward-only architecture: Kohonen then Grossberg layers
   • Training and operation phases
   • Applications and features
9. 7 Marks
   aWrite short notes on Adaptive Resonance Theory 1.
   How to answer

   • Need for Adaptive Resonance Theory
   • ART1 architecture: comparison and recognition fields
   • Vigilance parameter and reset mechanism
   • Stability-plasticity features
10. 7 Marks
    bDiscuss in brief the Fuzzy rules and reasoning.
    How to answer

    • Definition of fuzzy rule IF-THEN form
    • Types of fuzzy rules and reasoning methods
    • Generalized modus ponens/tollens
    • Example of fuzzy inference
11. 7 Marks
    aExplain in detail the block diagram of Fuzzy inference system.
    How to answer

    • Components: fuzzifier, knowledge base, inference engine, defuzzifier
    • Neat block diagram with signal flow
    • Function of each block in detail
    • Fuzzification to defuzzification process
12. 7 Marks
    bList and explain the applications of Fuzzy logic.
    How to answer

    • List of applications: control, appliances, automotive, medical
    • Explanation of 4-5 key applications
    • Advantages of fuzzy logic in applications
    • Examples with use-case
13. 7 Marks
    aDiscuss about the solution provided to travelling sales man problem using genetic algorithm.
    How to answer

    • TSP formulation as optimization problem
    • GA encoding, fitness, crossover, mutation for TSP
    • Selection and convergence to short tour
    • Example tour improvement steps
14. 7 Marks
    bWrite short notes on Ant Colony Optimization (ACO) and Particle Swarm Optimization (PSO).
    How to answer

    • ACO principle: pheromone, foraging behaviour
    • PSO principle: particles, pbest, gbest, velocity update
    • Algorithm steps for each
    • Comparison of applications
15. 14 Marks
    Write short notes on any two of the following:

    1. a)Hebb rule
    2. b)Back propagation network
    3. c)Convolution neural networks (CNN)
    4. d)Properties of Fuzzy sets
    How to answer

    • Select any two topics from four options
    • Definition, architecture/formula and working for each
    • Diagram or example where applicable
    • Applications/advantages conclusion