I am working on my masters in CS and I am starting my thesis. I think I've found the area I want to do it in which is using ML to make predictions in a certain sub-field. I have found no previous existing research on this and anything semi-related has been found in blogs and on the internet.
For those who have done a thesis ... what exactly is "good enough" for a thesis? How much is enough? I know I have to "add something new to my field" but again, how much? And what differentiates a thesis for a masters and a PhD?
I've tried asking my supervisor these questions and the response I basically get is, "Ill be able to say when it is enough". Fair enough, but I am hoping to get a bit more of a defined statement. Can a thesis be on a new application for machine learning that hasn't been done before? Again, how much is enough?
I am starting to work on this prediction of my sub-field using ML which my supervisor is allowing me to do though she is a bit hesitant as there is no research in it and she doesn't know the sub-field at all, she just knows the ML part. But there's no one else in the university with anything even closely related in stats, or econ, or math.
tl;dr Is a new application of Machine Learning worthy of a thesis? How much is "enough" for a thesis?
[link][4 comments]