Not too easy. Not too difficult.
Course Overview:
This course often feels like it's still in its early stages, so expect to feel like a "guinea pig." At the beginning of the semester, far too much homework was assigned, and completing it often required an absurd amount of Googling (this has since been somewhat addressed). Reading the textbook was not always helpful because the textbook provided contained typos and mathematical errors, which could be quite confusing. Grading took forever.
Course highlights:
The material on principal component analysis, classifiers (like support vector machines, etc.) and regression was probably the most interesting and most obviously practical information learned, along with perhaps Markov chains, hidden Markov models, etc.
Hours per week:
3-5 hours
Advice for students:
If you've taken AP Stats, expect the first few weeks to feel like very boring review. Homework assignments have proofs. The programming portions are coded in R, which will not technically be taught; expect to do a lot of Googling/self-learning/etc. Understanding mathematical notation is important; all the important formulas and theorems are presented rather formally in both the lecture and the textbook. Not knowing linear algebra will become very problematic by the last few weeks of the course, as concepts like eigenvalues/eigenvectors/etc. start showing up everywhere. Be patient; grades will probably take forever to be updated.