C-DAC: Evolutionary Computing - DST Sponsored SERC School 2009

About SERC
The Science and Engineering Research Council (SERC) is an apex body through which the Department of Science and Technology (DST), government of India, implements its programmes related to promoting and encouraging research in newly emerging/advanced and challenging areas of science and engineering.The SERC also sponsors state-of-the-art courses in engineering and science for the benefit of faculty of engineering colleges, as well as scientists and engineers working in industry, R & D organizations, CSIR laboratories, and Government Departments in the country. These courses offer an opportunity for increasing interaction between industry, academia and R & D organizations.

About the School
Artificial Intelligence and Machine learning are now being increasingly employed in several process industries including petroleum and petrochemical industries. These techniques are used for global optimization of reactors, unit operations and processes, for data driven modeling and several other tasks. These techniques enable the solution and implementation of large-scale real-life industrial problems, which were hitherto, not easily solvable.

The proposed SERC School aims to cover the fundamentals of a) evolutionary optimization b) classification and regression techniques employing novel machine learning paradigms like SVM and Random forests techniques for chemical engineers, using a tutorial approach all through.

A few case studies (based on recent research) on the optimization of industrial systems and on data driven modeling will be covered. In addition, application of these techniques to multi-objective optimization of a phthalic anhydride reactor, a heat exchanger network, and a pressure swing adsorption unit, online fault detection, dynamic optimization of reactors , data driven modeling of various processes will be covered. This should be useful for the participants in their teaching, and should also help them in undertaking research on challenging industrial problems.

Scope of the School
The school will introduce Machine learning and
Pattern Evolution algorithms:

1.The Machine learning methods would include
Support Vector Machines and Random Forests,
Kernel Methods.

2.The Pattern Evolution Methods would include
Genetic Algorithms (GA), Ant Colony System (ACS), Simulated Annealing (SA) and allied methods.

There would be pedagogic lectures to deal with each method along with important case studies for illustration. This will be well complemented with the hands-on sessions to demonstrate these techniques.

Case Studies
Apart from lectures there will be discussions on case studies, which include, but not restricted to, the following list:

Multi-objective optimization of a phthalic anhydride reactor
Optimization of a heat exchanger network
Optimization of a pressure swing adsorption unit
Online Fault Detection
Dynamic Optimization of Petrochemical & Bio-Reactors
Data driven modelling of various unit operations & processes
Identification of regimes of operations
QSPR

Faculty

Name	Affiliation	Topic
Prof. Santosh K. Gupta	IIT Bombay	SA, GA & Applications
Dr. Shrikant Bhat	ABB, Bangalore	SA, GA & Applications
Dr. Abhijit Kulkarni	SAS Sofwares, Pune	SVM Regression & Kernel Methods
Dr. Jayaraman V.K.	MIT and C-DAC, Pune	SVM, ACS and their applications
Dr. Sundararajan V.	C-DAC, Pune	Parallel GA

Who will Benefit
The school would be beneficial to:
Young faculty, Research Scientists in the Chemical, Petroleum or Petrochemical Engineering who are working in Academic Institutions and Government Research Laboratories and Practicing Scientists in the Industries.

Prerequisite
Graduation in Chemical, Petroleum or Petrochemical engineering or a higher degree with proven interest in research. At least few years of practicing experience in computer programming is essential.

School Organisation
The course will be for six days organized in eleven sessions. The morning sessions would be mainly lectures, which would give a perspective view of the area. Afternoon sessions would have hands-on and working group methods as well as presentation of case studies. More emphasis would be given on hands-on so that participants gain better understanding of the topics covered.

Course Material
Lecture notes and appropriate reference material would be given to all the participants.

School Fee
The proposed school is free for teachers from academic institutes and researchers from various Government aided research institutions.

Lunch and tea will be provided free of cost on all the course days. Applicants from academic institutions and Government funded laboratories must provide an official certificate from their institute. They will also be provided 3 tier A/C return rail fare by shortest route from the place of work on production of original tickets/receipts. Although the course is free for teachers from academia and scientists from Government funded R & D laboratories, a commitment fee of Rs. 2000/- must be paid in advance which shall be refunded upon attending the course.

Fee per participant from industry would be Rs. 6000/-

Registration
The interested participants are requested to fill in the registration form and send it along with a demand draft drawn in favour of "Centre For Development of Advanced Computing" payable at Pune towards the registration/commitment fee to the address given below. Due to limited number of available seats, registration would be done on the "first come first served" basis and fulfilling the prerequisites.

Download Registration Form

Deadline for Registration: 15 Nov 2009
Selected participants would be informed by 20 November 2009

Dates and Venue
The School will be held during 28th December to 2nd January 2010 at Maharashtra Institute of Technology, Pune. The participants are requested to report at MIT, Paud Road, Kothrud, Pune, Maharashtra 411038. Tel: 020 30273459

Contact Info:

Dr. V.K Jayraman
MIT Pune
C-DAC Pune
Tel: 020 – 25704228

Dr. V. Sundararajan
Scientific Engineering and Computing Group (SECG),
C-DAC, Pune University Campus,
Ganeshkhind Road,
Pune - 411007
Tel: 020 - 25704100
Fax: 020 – 25694004

Click here to send eMail