Serendip is an independent site partnering with faculty at multiple colleges and universities around the world. Happy exploring!

You are here

patterns and more patterns

julia_ferraioli's picture
Projects: 
I have always had problems with defining a term without using the term itself. Unpredictable order emerging out of chaos. While technically not the same meaning, it still confuses the issue. I was looking at several resources for entries on emergence, and of course my first choice was the entry in Wikipedia. It is a rather in-depth article, and an interesting read. In my research this summer, I was working with different machine learning algorithms that were dependent on user-generated constraints. The user would group data together by some sort of vague, unformed criteria, and the program would infer the user's intentions. The results were often quite surprising, as the user might not even be aware of how they were sorting information until the computer analyzed it. Unexpected patterns would arise. Not really sure if this qualifies as emergence, but as I was sitting in class on Monday, I kept thinking of that experience and the surprise that followed.

Comments

Doug Blank's picture

I'd say that it counts. But I do wonder about whether you were in fact "infering the user's intentions." This presupposes that the user has intentions---I'm a skeptic of the whole idea of intentionality. What if the users' preferences were themselves emergent phenomena? That doesn't change the outcome of your experiments, but maybe the way that you describe what you were actually modeling.
julia_ferraioli's picture

Looking back, I was in error in my description. However, sometimes the user would have a specific outcome in mind, and sometimes s/he wouldn't. What I was describing, I suppose, was what would happen if the user did not have a specific outcome in mind. Preferences being emergent phenomena? Would this be considering one user only, or many users working with the same data set?