group proposal feedback

[andrewpbray]

Thank you for providing three very different and creative topic proposals.

I find the first one to be the most promising This is both an important problem and pretty exciting given the existence of this data set. The main challenge will be thinking through how to translate text-as-data to useful features for building the classifier.

The second topic is rather vague as stated - what exactly is the response? What is the unit of observation? I think there is a good project here, but it would likely utilize spatial models that are beyond the scope of this course.

The main challenge with the final one is that you'd be working with country-level data, which difficult to work with because there are only so many countries and you're averaging over a ton of variability when you aggregate up to the country level.