I'm an ML fan. Unfortunately, when I tell people I know ML, I basically mean:
1) I've done a Coursera course and an ML course at uni
2) I've done a project or two on classification/clustering
3) I've read one paper on SVM optimizations(and that was really hard to do) and glanced at the abstracts of hundreds.
In other words, I'm great at applying the things I learn about ML from the internet, but when I open up last year's NIPS or ICMLC proceedings, I don't understand much.
I now have the daunting task of finding a professor willing to host me for 6 months for my Bachelor thesis.
Who in their right mind would want to host someone who just knows the basic math behind some ML algorithms that he picked up from Coursera?
My point is, I'm not sure how to approach the process of searching for a professor, because my interest in and understanding of ML isn't evolved enough to look at the publications of a professor, and say "Hey, this guy's interests match up perfectly with mine! This might work!"
I've been told to spam mails to everyone in an ML research group, mention something about one of their papers, and ask nicely to be part of the family. But
A) There is no chance I'd be able to do justice to claiming I read their papers, so short of listing out my skills, and experience with ML, I can't think of anything else I can do.
B) Sending mails to everyone like that is just plain insulting
So my questions are:
1) As a professor, what are you looking for in a student who is REALLY REALLY interested in doing a Bachelor thesis under you/with your ML research group?
2) As an undergraduate student who's at my level of ML exposure, what should I be doing right now (other than reading a textbook on ML, I'm trying to do that anyway) with the motive of securing a Bachelor Thesis?
3) How do I find these elusive, magical professors who are willing to take me in?
PS: My university professors won't help me at all.
[link][3 comments]