- 4 reviews
- 4 completed
I consider this as the best MOOC I have taken, among some 15 or so completed. I had previously taken courses that used R and was eager for more practice. Your background will determine how difficult you think this is. Some other MOOCs had introduced even more challenging concepts (Stanford's excellent Introduction to Statistical Learning and Andrew Ng's Machine Learning course on Coursera). Neither of those offered nearly as many problems. If you think of quizzes and homework as puzzles, the volume of puzzles in this course was quite gratifying. Indeed, the sheer quantity of problems in this one made the experience of taking it qualitatively different from other MOOCs on similar topics. I really feel like this course helped me master several concepts and techniques I had previously seen. The required kaggle competition was a fun interlude, and might well have served as a gateway to a future addiction. The final two weeks entailed optimizing using Excel Solver. I was already quite facile with Solver, but the course still managed to teach me a few things in the realm of integer optimization. The MITx team did a great job using all the features of the edX platform to their advantage. I especially appreciated having R scripts and Excel workbooks demonstrating ALL the lecture examples.
The other reviewers' criticisms are valid. It's short, easy, and you can argue its volume of content doesn't amount to a whole course. EXCEPT... I had known about git and github for years and had never used either. Same thing with markdown. The benefit of this course to me was forcing me to use these things, which I didn't appreciate until I had tried them. I found this course much more useful in retrospect, now that I have moved on to subsequent Data Science courses where some of these tools make the course much easier to navigate. As for the minimal volume of time, I suppose for an absolute beginner to R it is a baby step toward digging in to the later courses. I might suggest to the JH profs they include this as a free course and not require it as part of the specialization.
I took the pre-Udacity version, Fall of 2011. Previous experience was nil. The experience led me since then to take another 20 or so MOOCs, of which I have completed about 15. Taking this was a lot of fun.
Prior experience: whatever Sebastian Thrun introduced in his pre-Udacity AI course. This course taught a lot of machine learning techniques, and spent a good deal of time discussing when and why to use one or another. The course I took was taught in Octave, which was easy to pick up.