Soft Computing (CSA301)

📘 Syllabus (CSA301 - Theory)

• Unit 1: Introduction to Soft Computing
  • A. Concept of computing systems. What is Soft Computing?
  • B. "Soft" Computing versus "Hard" Computing.
  • C. Characteristics of Soft Computing, Some applications of Soft Computing techniques.
• Unit 2: Fuzzy Logic
  • A. Introduction to Fuzzy logic, Fuzzy sets and membership functions.
  • B. Operations on Fuzzy sets, Fuzzy relations, rules, propositions, implications and inferences.
  • C. Defuzzification techniques, Fuzzy logic controller design, Some real life societal applications of Fuzzy logic.
• Unit 3: Fuzzy Inference System
  • A. Fuzzy Inference Systems, Different Fuzzy Models: Mamdani Fuzzy Models, Sugeno Fuzzy Models.
  • B. Tsukamoto Fuzzy Models, Input Space Partitioning and Fuzzy Modeling.
  • C. Neuro Fuzzy Modelling: Adaptive Neuro-Fuzzy Inference Systems (ANFIS), Architecture, Hybrid Learning Algorithm, Learning Method that Cross-fertilize ANFIS and RBFN.
• Unit 4: Swarm and Evolutionary Algorithms
  • A. Concept of "Genetics" and "Evolution" and its application to probabilistic search techniques.
  • B. Basic GA framework and different GA architectures, GA operators: Encoding, Crossover, Selection, Mutation, Solving single-objective optimization problems.
  • C. Swarm Optimization: Introduction to Ant Colony Optimization, Particle Swarm Optimization etc.
• Unit 5: Multi-objective Optimization Problem Solving
  • A. Concept of multi-objective optimization problems (MOOPs) and issues of solving them.
  • B. Multi-Objective Evolutionary Algorithm (MOEA), Non-Pareto approaches to solve MOOPs, Pareto-based approaches to solve MOOPs, Some applications with MOEAs.

🧑‍💻 Self-Learning (SL) Activities

As per NEP 2020 guidelines, this course carries student-directed self-learning hours (L-T-P-SL: 3-0-0-3).

• 1. Problem Solving / Coding using MATLAB or Python
  • Implement core SC algorithms: Fuzzy Logic Controller design (skfuzzy), GA (DEAP), PSO (pyswarms), NSGA-II (pymoo) on real or benchmark datasets.
  • Evaluation: Based on coding solution submitted — correctness, algorithm performance, code quality and documentation of results.
• 2. Technical Video-Based Learning
  • NPTEL 'Soft Computing' course (IIT Kharagpur) or equivalent — focused on FIS, GA and MOEAs as per units covered in class.
  • Evaluation: Report/presentation based on video learning outcomes — concept clarity, diagrams and Q&A performance during viva-voce.
• 3. Assignment Writing (Mathematical & Numerical)
  • Derivations and worked numericals: fuzzification & defuzzification, Mamdani FIS step-by-step, GA crossover/mutation trace, PSO velocity update, Pareto dominance ranking.
  • Evaluation: Correctness of derivations, step-by-step working and coverage of assigned problem statements.
• 4. Real-World Case Study
  • Study a published papers and review application of soft computing (e.g., fuzzy medical diagnosis, GA-based network routing, PSO for power systems) and prepare an analytical report and present in the class.
  • Evaluation: Depth of analysis, technical accuracy, referencing quality and presentation of findings.

📚 Lectures

Theory lecture materials have been uploaded unit-wise.