MCA-237
Soft Computing
Computing that tolerates imprecision — fuzzy sets and fuzzy logic, genetic algorithms, neural networks, and the hybrids built from them.
- Semester
- Semester 3
- Credits
- 3
- Scheme
- 3 lectures a week
- Contact hours
- 40 across four units
- Category
- Core Elective - III
What the paper is
Soft Computing is defined by contrast: hard computing demands precision and gives an exact answer, soft computing accepts imprecision and gives a good enough one quickly. The paper covers the three main techniques that do this — fuzzy logic, evolutionary computation and neural networks — and then combines them.
Fuzzy logic gets two units because it needs them. Unit I builds fuzzy sets against classical sets and membership functions; Unit II does the operations, fuzzy arithmetic, rule generation, defuzzification and controller design. The fuzzy logic controller at the end is what the whole thread was for.
Units III and IV are genetic algorithms and neural networks respectively, and the last section on hybrid systems — neuro-fuzzy, GA for network design, GA for optimisation — is where the paper makes its argument that these three are complementary rather than competing.
What to concentrate on
- Fuzzy set operations and defuzzification by the centroid method are the most reliable numerical questions. Work a complete example: fuzzify, apply rules, aggregate, defuzzify.
- For genetic algorithms, be able to run two or three generations by hand on a small problem — selection, crossover, mutation, fitness. The trace is the question.
- Backpropagation on a tiny network with one hidden layer is the neural network equivalent. Doing one forward and one backward pass by hand makes the rest of Unit IV read easily.
Syllabus, unit by unit
Summarised from the approved scheme. Hours are the contact hours the scheme allots, which is a fair guide to where the marks sit.
Unit I · Introduction and fuzzy sets
10 hours- Soft computing concepts; the importance of tolerance for imprecision and uncertainty
- Soft versus hard computing; characteristics of soft computing
- Applications of soft computing techniques
- Introduction to fuzzy sets, overview of classical sets, membership functions
Unit II · Fuzzy operations, arithmetic and control
10 hours- Operations on fuzzy sets: complement, intersection, union, combinations and aggregation operations
- Fuzzy rule generation, fuzzy propositions, implications and inferences
- Fuzzy arithmetic: fuzzy numbers, linguistic variables, arithmetic on intervals and numbers, the lattice of fuzzy numbers, fuzzy equations
- Defuzzification techniques, fuzzy logic controller design and applications
Unit III · Evolutionary computation
10 hours- Genetic algorithms and genetic programming
- The basic GA framework and different GA architectures
- Operators of genetic algorithms
- Analysis of selection operations and convergence of genetic algorithms
Unit IV · Neural networks and hybrid systems
10 hours- Biological neurons and their simulation for problem solving
- Introduction to artificial neural networks and their architectures
- Training techniques: perceptrons, training rules, the backpropagation algorithm
- Multilayer perceptron model, Hopfield networks, applications to real-life problems
- Hybrid systems: neuro-fuzzy systems, genetic algorithms for neural network design and learning, fuzzy logic and genetic algorithms for optimisation
How it is examined
- Question 1 is compulsory, covers the whole syllabus, and is made of objective or short-answer questions. It carries 12 marks.
- The rest of the paper is four units matching the four above. Each unit sets two questions and you attempt one, so there are four more answers at 12 marks each.
- Examiners are asked to map every question to a course outcome using Bloom's taxonomy, which is why questions are phrased as “apply”, “analyse” or “evaluate” rather than only “explain”.
The pass mark is 40 out of 100 in every paper.
Learning objectives
- Fuzzy logic and its applications
- Artificial neural networks and their applications
- Applications of soft computing across a variety of domains
Assumed beforehand
- A strong mathematical background
- Proficiency with algorithms
- Programming skills in C, C++, Java or MATLAB
- Critical thinking and problem solving skills
Paired in the scheme with MCA-267 Lab based on Core Elective III.
Books it is taught from
- TB1J. A. Anderson, “An Introduction to Neural Networks”, PHI, 1999
- TB2G. J. Klir and B. Yuan, “Fuzzy Sets and Fuzzy Logic”, PHI, 1995
- TB3S. N. Sivanandam and S. N. Deepa, “Principles of Soft Computing”, Wiley India, 2008
- TB4Samir Roy and Udit Chakraborty, “Introduction to Soft Computing: Neuro-Fuzzy and Genetic Algorithms”, Pearson Education
Notes for MCA-237
Lecture notes, question papers and assignments filed against MCA-237 are in the library. Everything there has been checked by a moderator before it appears, and an account is free — new ones are approved by hand, which is what keeps the shelf clean.