SCHOOL OF COMPUTER SCIENCE & ENGINEERING, ITER, BBSR-30
LESSON PLAN Course Information Session: Jan-2014 to Apr-2014
1. Course Number and Name: EC 452, Soft Computing 2. Credits and Contact Hours: 3, 3-0-0 3. Target Student: B. Tech, 8 th Semester 4. Instructors Name: Dr. Debahuti Mishra (Coordinator) Contact # debahutimishra@soauniversity.ac.in Mr. Chinmaya Kumar Swain Contact # chinmayaswain@soauniversity.ac.in Mrs. Swagatika Devi Contact# swagatikadevi@soauniversity.ac.in Mrs. Shruti Mishra Contact# shrutimishra@soauniversity.ac.in 5. Text Book and References: Text Book: 1. S. Haykin, Neural Networks A Comprehensive Foundation, Pearson Education, India 2. Martin T. Hagan, Howard B. Demuth, Mark H. Beale; Neural Network Design; (ISBN: 0-9717321-0-8); Thomson 2002 3. Jang, Sun and Mizutani; Neuro-Fuzzy and Soft-Computing A computational approach to learning and machine intelligence, Prentice Hall of India 4. David E. Goldberg, Genetic Algorithms in search, Optimization and machine learning, 1989. Reference Book: 1. Satish Kumar, A Classroom Approach, Neural Networks: Tata McGraw Hill, 2004, ISBN: 9780070482920
6. Specific Course Information: a. Course Description: The course deals with the foundations of some paradigms that are relevant for the treatment of uncertainty in computing: using soft computing approaches such as neural networks, fuzzy logic and evolutionary computation (GA and SA). The main goal is to achieve a thorough understanding of the foundations of these paradigms, and also to experiment with their use and practical application in a number of problems, especially in the field of knowledge representation and machine learning.
b. Prerequisites and/or Co-prerequisites
Prerequisites Knowledge of biological process of human brain, a solid mathematical and logical background is essential. Basically the knowledge of calculus, algebra, and programming, algorithms and data structures. Co-prerequisites Any numerical computational package, basically MATLAB/ Scilab/ LABView.
7. Specific goals for the course:
At the end of the semester the students will be able to: Acquire the basic skills to solve real problems using soft computing techniques. Integrate and make autonomous use of their knowledge. Deal with complex phenomena, issues and situations, and develop potential for professional activities that demand considerable autonomy, or for research and development work. Will know the problems that are still not resolved in various fields of computation, where soft computing approaches can therefore be used, as they are open research fields. 8. Brief list of topics to be covered: Contact hour Topics to be covered Remarks (if any) Week #1 MODULE I 1. AI and soft computing (SC), From conventional AI to Computational Intelligence, Main goals of SC, SC constituents (Neural network, Fuzzy and evolutionary computing)
2. Basic concepts of neural network (NN), benefits of NN, human brain, models of a neuron.
3. Types of activation functions (Threshold, Piecewise linear, Sigmoid)
Week #2 4. NN as directed graph, Feedback, NN architectures (single layer, multi layer and recurrent network)
23. Basic concepts of genetic algorithm (GA), Search space, working principle. MODULE II 24. Encoding: Binary, Octal, Decoding Week #9 25. Fitness function, Selection: Roulette-wheel 26. Tournament, Rank and steady state, Elitism 27. Crossover: Single, two-point and multi-point cross over, Uniform, matrix and cross over rate.
Week #10 28. Inversion and deletion: Inversion, deletion and duplication, deletion and regeneration.
29. Segregation, cross over and inversion, mutation, mutation rate.
30. Ant colony optimization, Ant foraging behaviour Week #11 31. Combinatorial optimization, Routing in communication network
32. Applications: Particle swarm optimization (PSO): Introduction and working principle MODULE III 33. Application of PSO in TSP problem Week #12 33. Bacteria foraging optimization (BFO) introduction, working principle
34. Application of PSO in pattern classification 35. Application of BFO in pattern classification
Evaluation in each of the above component (except the "attendance" component) will be based on the following way:
i. Understanding of the question: 30% ii. Response style and formatting: 30% iii. Final successful response: 20% iv. Giving response in the first attempt: 20%
There may be "n" number of quizzes and the best response will be taken into consideration for the final mark in quiz. We often include questions on the quizzes and final that can be answered only by faithful lecture attendance and response to assignments.