I'm a professor analyzing data of university student attributes and performance. I'm using it to try to predict/identify students who will likely have problems graduating. It's 10 years of data from a large university in the U.S. Lots of features. It should be very interesting.
The tricky part is that I can't share the data with you, so we'd have to do some sort of live coding together. I've been teaching myself python, so I'd like to use that software. So I'd be using you as both a consultant for the modelling but also a mentor to help me become a better python programming.
For the first part of this project, I'm envisioning about 10 hours of work together, though I'd like to start with just an hour or 2 to see how we work together. After that 10 hours, I would probably continue to hire you periodically for consulting.
As I said, the data (which is already pretty clean) is 10 years of data which includes student attributes (GPA, demographics, SAT scores, performance in individual classes, graduation GPA, major...). The first part of the project would be data exploration and visualization. The second part would be statistical modeling. For instance, I'd like to model things like the probability of successfully graduating, the probability of switching majors, the probability of a significant decrease in GPA during the, say, next semester...
What I'd like to do is have a preliminary meeting with the freelancer and try to work together for an hour or so, and if it works well then we can continue. I'm envisioning about 10 hours of work over the next month with the possibility of increasing that depending on my needs.
The freelancer should have significant experience with python and scientific libraries. I'd prefer someone with strong statistical/machine learning skills too.