DBATU · Final Year B.Tech Electronics & Telecommunication Engineering · Non-NEP
Soft Computing
BTETOE704A · Elective III · Semester VII · 4 credits · Continuous assessment 20 + mid-semester 20 + end-semester 60
One of the Elective III options.
Unit 1: Artificial Neural Networks I
7 hoursBiological and artificial neurons, bias and threshold; the McCulloch–Pitts model for AND, OR and XOR; topologies and learning paradigms; the linear neuron, error energy and gradient descent; activation functions; Hebbian and delta rules; the perceptron and its limits.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Biological and artificial neurons
- McCulloch-Pitts neuron model
- Learning rules: supervised, unsupervised and reinforcement
- Activation functions
- Single-layer perceptron
- Linear separability and the XOR problem
Unit 2: Artificial Neural Networks II
7 hoursThe multilayer perceptron and backpropagation for classification and regression; self-organising maps and k-means; learning vector quantisation; RBF networks and Cover's theorem; Hopfield networks and associative memories.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Multilayer perceptron
- Error backpropagation
- Kohonen self-organising maps
- Learning vector quantisation
- Radial basis function networks
- Hopfield networks and associative memory
Unit 3: Fuzzy Logic I
7 hoursFuzzy numbers and sets, continuous and discrete; operations on fuzzy sets; membership functions (core, boundary, support); linguistic terms; fuzzy relations and T-norm and T-conorm composition; fuzzy if-then rules.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Fuzzy sets and operations
- Membership functions and fuzzification
- Fuzzy relations and composition
- Fuzzy if-then rules
Unit 4: Fuzzy Logic II
7 hoursFuzzification and membership assignment; max-membership, centroid and weighted-average defuzzification; implication rules; Mamdani, Sugeno and Tsukamoto inference systems; a two-input Mamdani FIS.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Defuzzification methods
- Fuzzy inference systems
Unit 5: Fuzzy Control Systems and ANFIS
7 hoursFuzzy control design and decision surfaces; fuzzy logic controllers compared with PID; the Mamdani FLC architecture and the aircraft-landing example; ANFIS architecture and hybrid learning, its limits, and ANFIS/CANFIS for regression.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Fuzzy logic controller design
- ANFIS: adaptive neuro-fuzzy inference systems
Unit contents are summarised from the official DBATU syllabus . AiCanCode.org is not affiliated with Dr. Babasaheb Ambedkar Technological University (DBATU).