IT-701 Soft Computing - Dec 2025 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
   aExplain the structure and functioning of a "biological neural network" and compare it with an "Artificial Neural Network (ANN)".
   How to answer

   • Define biological neuron components: dendrites, soma, axon, synapses
   • Explain signal transmission and firing/threshold functioning
   • Define ANN model: inputs, weights, summation, activation, output
   • Compare on structure, processing, learning, fault tolerance with table/diagram
2. 7 Marks
   bDescribe the "McCulloch-Pitts neuron model" and explain how it helps in solving simple logic functions.
   How to answer

   • State McCulloch-Pitts model assumptions and binary threshold logic
   • Explain weighted sum and threshold activation with formula
   • Show construction for AND, OR, NOT logic functions
   • Discuss limitations and significance
3. 7 Marks
   aDefine "Linear Separability" and explain why the "XOR problem" cannot be solved using a single-layer perceptron?
   How to answer

   • Define linear separability with decision boundary concept
   • Explain XOR truth table and plot showing non-separability
   • Prove no single line can separate XOR classes
   • Conclude need for multilayer/hidden layer
4. 7 Marks
   bDiscuss the "Perceptron Learning Rule" and its convergence characteristics.
   How to answer

   • State perceptron learning rule weight update formula
   • Explain error-correction training steps
   • State perceptron convergence theorem conditions
   • Discuss convergence for linearly separable patterns and limitations
5. 7 Marks
   aCompare "Adaline" and "Madaline" networks with respect to their architecture and learning algorithms.
   How to answer

   • Describe Adaline architecture with single linear unit and LMS
   • Describe Madaline architecture with multiple Adalines and majority vote
   • Compare learning algorithms: LMS vs MRII/Madaline rule
   • Summarize differences in table
6. 7 Marks
   bExplain the "Back Propagation Algorithm" and discuss how it is used for training multilayer neural networks.
   How to answer

   • Outline forward pass and error back-propagation concept
   • Give weight update with gradient descent and chain rule steps
   • Explain training phases for multilayer networks
   • Mention advantages and issues like vanishing gradient
7. 7 Marks
   aDescribe the "Kohonen Self Organizing Map (SOM)" and explain its working process with a neat diagram.
   How to answer

   • Define SOM as competitive unsupervised network
   • Explain competition, cooperation and adaptation steps
   • Provide neat labelled architecture/topological map diagram
   • Mention applications in clustering and visualization
8. 7 Marks
   bDifferentiate between "Convolutional Neural Networks (CNN)" and "Recurrent Neural Networks (RNN)" with examples of their applications.
   How to answer

   • Define CNN structure for spatial data vs RNN with recurrent loops for sequential data
   • Differentiate on architecture, memory, training
   • Give CNN examples: image classification; RNN examples: NLP, time-series
   • Summarize in comparison table
9. 7 Marks
   aExplain the "properties of fuzzy sets" and their "set-theoretic operations" with examples.
   How to answer

   • Define fuzzy set, membership function and key properties
   • Explain union, intersection, complement operations with formulas
   • Give numerical/linguistic examples for each operation
   • Mention properties like commutativity, distributivity
10. 7 Marks
    bConstruct a "Fuzzy Inference System (FIS)" for temperature control using appropriate "fuzzy rules".
    How to answer

    • Outline FIS blocks: fuzzification, rule base, inference, defuzzification
    • Define linguistic variables and fuzzy rules for temperature control
    • Show example rule set and inference flow
    • Explain defuzzified control output
11. 7 Marks
    aExplain "fuzzy if-then rules" with an example. How does fuzzy reasoning differ from conventional logic reasoning?
    How to answer

    • Define fuzzy if-then rule structure with antecedent-consequent
    • Give example e.g. IF temperature is high THEN fan speed is high
    • Explain Mamdani/Sugeno reasoning steps
    • Contrast with crisp Boolean logic on truth values and boundaries
12. 7 Marks
    bDiscuss the "applications of fuzzy logic" in solving engineering problems.
    How to answer

    • Define role of fuzzy logic in handling uncertainty
    • List engineering domains: control, automotive, appliances, decision systems
    • Give specific application examples
    • Discuss advantages in real-world problems
13. 7 Marks
    aExplain the Genetic Algorithm (GA) with the help of a neat flowchart and discuss its main operators: selection, crossover, and mutation.
    How to answer

    • Outline GA cycle: initialization, fitness, selection, crossover, mutation
    • Provide neat flowchart with termination condition
    • Explain each operator with example
    • Mention fitness function and elitism
14. 7 Marks
    bDescribe the Schema Theorem and explain how it supports the working of GA?
    How to answer

    • Define schema, order, defining length and fitness concepts
    • State Schema Theorem inequality/formula
    • Explain growth of short, low-order, above-average schemata
    • Link to implicit parallelism supporting GA search
15. 14 Marks
    Write a short note on any two.

    1. a)Ant Colony Optimization (ACO)
    2. b)Particle Swarm Optimization (PSO)
    3. c)Counter Propagation Networks (CPN)
    4. d)Hebb Learning Rule
    How to answer

    • For each chosen topic give definition and principle
    • Explain working steps/algorithm with diagram/formula
    • Mention key parameters and applications
    • Conclude with advantages/limitations