Pattern recognition: neither deduction nor induction

John Wilkins, Evolving Thoughts, Nov 19, 2014
Commentary by Stephen Downes

I've spoken many times about the idea that to know is to recognize and yet I've rarely (if ever) followed it up with a reference. Part of the reason is that I'm lazy, and part of the reason is that I've slowly developed this idea over time. Still. It's not like I'm alone here. So we have this article by John Wilkins making the distinction between pattern recognition and traditional epistemology (which views knowledge as a type of deductive or inductive inference). I don't see pattern recognition as a means of classification so much; rather, I see recognition as a process that stimulates memories directly, without the need for the mechanism (and language) of classification. A lot has been written on pattern recognition and I think we should take it seriously as a way of representing knowledge tasks as types of direct perception rather than as inferential or encoding processes.

Views: 0 today, 179 total (since January 1, 2017).[Direct Link]