This question paper belongs to the subject Soft Computing, designed for undergraduate engineering students. The paper carries a total of 100 marks and is divided into three sections: A, B, and C. Students must attempt all sections, and any missing data may be assumed suitably.
The paper evaluates a student's conceptual understanding, analytical ability, and problem-solving skills in the domains of Neural Networks, Fuzzy Logic, Genetic Algorithms, and Hybrid Soft Computing Models.
Section A (20 Marks – Short Answer Questions)
This section contains 10 brief questions, each carrying 2 marks.
Questions test fundamental knowledge of:
Learning rate
Soft computing concepts
Basics of genetics (cell & chromosomes)
Fuzzy vs crisp logic
Fuzzy Cartesian product
Disjunctive sum of fuzzy sets
Bias function in neural networks
Complexity
Activation functions
Membership functions
This section checks the student’s basic theoretical foundation.
Section B (30 Marks – Descriptive Questions)
Students must attempt any 3 out of 5 questions, each worth 10 marks.
This section requires detailed explanations of:
Realization of AND using perceptron
Characteristics & applications of neural networks
Multi-layer perceptron & soft computing applications
Neuro-fuzzy systems
Fuzzy automata
These questions examine the student’s ability to explain concepts with clarity and depth.
Section C (50 Marks – Long Answer / Numerical / Analytical Problems)
This section contains five sub-sections, and the student must attempt one part from each, making it the most scoring section.
It includes:
Numerical problems (e.g., KSOM clustering, back-propagation iteration)
Conceptual explanations (e.g., fuzzy control, hybrid systems)
Detailed algorithms (e.g., genetic algorithm phases)
System models (e.g., ANFIS with Mamdani model)
Learning techniques (supervised vs unsupervised learning)
This section assesses the student's analytical skills, problem-solving ability, and application of soft computing methods.
Purpose of the Question Paper
The paper is designed to test:
Conceptual clarity in soft computing
Understanding of fuzzy logic, ANN, genetic algorithms
Ability to perform numerical computations
Skill in designing intelligent hybrid systems
Practical knowledge of machine learning fundamentals
Ideal For
B.Tech/B.E. – Computer Science, IT, AI/ML
MCA / M.Sc. students
Anyone preparing for competitive exams involving Soft Computing
Students needing structured practice for neural network & fuzzy logic problems
