My grand plan for my masters degree was to study data mining, but more and more I'm realising that maybe I didn't really know what data mining is all about when I decided that. Now, as someone who has no intention of ever doing phD, and I say that with 99% certainty, I think I am going to change my research area. UIUC has THE professor of data mining field (well, he basically created the field) and fortunately I am doing a small project with him and his phD students, so it would be just supa if I could build on that project to write my thesis... But from what I have learnt in the data mining course so far, everything seems to be very.... textbooky. Great that someone came up with some new algorithm to materialise data cube faster or mine some patterns faster (by THE professor and his students 90% of the time which is actually very impressive), but from what I can see they are all within the context of experimental amount of data, with optimistic restrictions, mainly for the purpose of proving the algorithm to publish a conference paper. I think it is my "industry background" that is kinda dictating how I accept things that are happening in academia... I know data mining and algorithms are being used in the industry as we speak, but for some reason I am not really convinced. Maybe if I see a real world implementation of it, I might change my mind but in the mean time I'm heading towards database systems and data security/privacy area... something that I can see the real use in the industry.
No comments:
Post a Comment