Start
Nov 23, 2016 - 16:00
End
Nov 23, 2016 - 17:00
Venue
Rm. No. 240 Chemical Engg. Dept.
Event Type
Speaker
Shankar Narasimhan Chair Professor in the Department of Chemical Engineering at IIT Madras.
Title
Colours and Flavours of PCA.
Abstract: Principal Components Analysis (PCA) is a multivariate data analysis technique which was first introduced by Pearson in 1901 although the nomenclature was provided by Hotelling in 1933. This technique has gained significant popularity in recent years due to the tremendous explosion in the volume of data that is being generated. It is mainly used as a data compression technique which can aid in better visualization and help reduce storage. Since its humble beginnings as a three letter acronym it has evolved by acquiring another alphabet (FPCA DPCA PPCA SPCA CPCA IPCA KPCA) or two (MLPCA SWPCA). Its extended family includes recent members such as independent components analysis (ICA) and network component analysis (NCA). Other members of this family (which do not bear the same family name) include non-negative matrix factorization (NMF) and factor analysis (FA). This talk will provide a taxonomy of all these methods using a measurement error modelling framework. The basic assumptions of this model for deriving PCA will be elucidated and all other variants will be derived by modifying the assumptions regarding either (a) the noise corrupting the measurements (b) the model structure relating the different variables or (c) the model structure that relates the different samples. It is hoped that this taxonomy will provide a better theoretical understanding of the different methods resulting in their proper use in data analysis. Scope for adding new meaningful acronyms to this alphabet soup will also be pointed out.Bio: Shankar Narasimhan is currently the M.S. Ananth Institute Chair Professor in the Department of Chemical Engineering at IIT Madras. He obtained his Bachelor’s degree from IIT Madras in 1982 and his MS and PhD degrees from Northwestern University Illinois USA in 1984 and 1987 respectively. Prior to joining IIT Madras he was an Associate Professor in chemical engineering department at IIT Kanpur. Shankar Narasimhan’s major research interests are in the areas of Data Mining Process Design and Optimization Fault Detection and Diagnosis (FDD) and Fault Tolerant Control. Shankar Narasimhan is well known for his work in the area of Data Reconciliation in which he has co-authored several important papers and a book which has received critical appreciation in India and abroad. Three of his publications in AIChE Journal were included in a select list of 210 papers as having made a significant contribution to Process System Engineering. Along with Engineers India Ltd. Gurgaon he has developed a software package for applying data reconciliation and gross error detection technique to process industries. For this work he received the Engineers India Research Award instituted at IIT Kanpur in 1994-1995. In collaboration with a colleague from Civil Engineering he has also developed a software package for leak detection in gas pipeline networks which has been field tested on a pipeline owned and operated by GAIL. In the area of process design he has made important and seminal contributions in the design of sensor networks heat exchanger networks and water distribution networks. Shankar Narasimhan has been a visiting professor at the Centre for Automatic Control in Nancy France Purdue University Clarkson University and Texas Tech University in USA and the University of Alberta in Canada. He has also spent summer internships at Engineers India Ltd. R&D Centre in Gurgaon Honeywell Technology Solutions Ltd. R&D Centre at Bangalore and ABB Global Services Ltd. Bangalore as part of high-level industry-academia interactions. Along with Prof. Raghunathan Rengaswamy of IIT Madras he co-founded Gyan Data Pvt. Ltd. in 2011 which specializes in providing services and products in data analytics process modeling and optimization. Shankar Narasimhan was elected as a Fellow of the Indian National Academy of Engineering in 2012.