Encounters with Fuzziness and Ambiguity in Patterns – A Memorable Journey

Abstract

I joined the Indian Statistical Institute (ISI), Calcutta on March 01, 1975 to work on a PhD in the area of pattern recognition and man-machine communication by voice as a CSIR Senior Research Fellow of the Government of India. I had no idea about pattern recognition and neither were there any text books on this subject; only a few edited volumes, mostly by Prof. K.S. Fu, were available in our library or in the market. From these, I started to pick up the basics of sequential pattern recognition using statistical approaches. One day my thesis advisor, Prof. D. Dutta Majumder, gaveme a typewritten note, of about four pages titled something like “Pattern classification with property sets”, written by Prof. Ramesh Jain who had presented the concept in a seminar in Calcutta. From it, I got some idea of the fuzzy properties of a pattern and the concept of its multi-class belonging based on them - which appealed to me very much as it seemed to be very natural for decision-making in real-life problems. There, I also got the reference of Prof. Zadeh’s famous paper, “Outline of a New Approach to the Analysis of Complex Systems and Decision Processes”, which had appeared in the IEEE Transactions on SMC (vol. 3, pp. 28–44), in 1973.