Rahul Kala

Computational Intelligence

Mesmerized by the storm of increase of intelligence in the machines and the storm of decrease of intelligence in the humans, the institute decided to start a flagship course that teaches the students everything about intelligent machines in a single course, and this course was born. Studied by students from different academic coursework backgrounds, with different projects in the past, and with different requirements - this mandatory course is the modern day equivalent of unity in diversity. The course sometimes covers topics that all but one already know, and sometimes leaves topics that none know.

Like many other courses at this institute, a good part of the course is on neural networks and machine learning with the difference that for some post-graduate students this is the first such offering. The course slowly dissects the commonly seen intelligent systems, takes the technology outside for examination, notes the basic constituents, and teaches the generics. Because machine learning is a hype, the course covers different applications, problems, and some tools. The leftover time goes into evolutionary computation and fuzzy logic. To make the instructor look intellectual, the complexity is added by discussions on different hybrid systems involving multiple techniques at once. The laboratory part is easy to solve given a wide array of libraries available. The course looks into some machine learning data sets, regression and classification problems, and some computer vision problems to gets a hands-on and let students brag about what their codes do in front of non-IT friends and family.

Syllabus

Machine Learning Primer
Application Primer: Vision, Feature Engineering
Classification Primer
Neural Networks
Support Vector Machines
Principle Component Analysis
Self Organizing Maps
Recurrent Neural Networks
Committee Machines
Genetic Algorithms
Fuzzy Logic
Hybrid Systems: ANFIS, Evolutionary Neural Networks, Feature Selection using Genetic Algorithm, Evolutionary Fuzzy Systems

Lab Syllabus

Regression problems
Object recognition problem
Committee Machines

Pre-Requisites

The course gets the highest possible diversity of students, so there is no means to implement a pre-requisite. It could be a cakewalk if you've doen courses or projects related to soft computing, machine learning, deep learning, etc.

Text/Reference Material

• Main Text Book 1: S. Haykin (2009) Neural Networks and Learning Machines, Pearson, Upper Saddle River, NJ.
• Main Text Book 2: T.J. Ross (2010) Fuzzy Logic with Engineering Applications, Wiley, West Sussex, UK.
• Main Text Book 3: M. Mitchell (1998) An Introduction to Genetic Algorithms, MIT Press, Cambridge, MA.
• Main Text Book 4: L. I. Kuncheva (2004) Combining Pattern Classifiers Methods and Algorithms, Wiley, Hoboken, NJ.
• Additional papers are taken for supplementary topics.