ML Foundations: a Review

I just today finished the first course of six in the University of Washington's Coursera class on Machine Learning. Here are my thoughts so far:

• In comparison to the algorithms class that I'm auditing (also a Coursera class), this course has been a breeze. There's probably 60-90 minutes of video per week. I wanted to go slowly though, to let things sink in. I took my time and I still finished the 6-week course in about 3 weeks. I trust that the other courses in the series will be a lot more in-depth
• The last week, on deep learning, felt a little rushed to me, but everything else was easy to follow.
• I'm not a fan of the GraphLab tool they use. I understand the choice and usage though (besides the affiliation of one of the lecturers)--In terms of data exploration, it is easy to get started, no hiccups or learning ways around a package, and it seems intuitive and responsive. But it'd be nice if there was some intellisense to list parameters for all of the functions mentioned. Also, the awkward syntax (e.g. for filtering) was a huge turn off
• I felt that a lot of time being spent waiting for the presenter to finish writing or typing
• Despite all of these negative things said, this survey course has made me excited to learn ML!! I wanted to get into Computer Vision a couple of years ago (it was one of my favorite classes in college), but I ended up spending too much time on studying for the GRE and learning digital image processing before deciding to commit my time to functional programming. I now see a potential path to get back into CV, by first also becoming interested in ML.

Side note, I haven't worked with Python for about 5 years. Is it just my imagination or has it completely exploded in popularity while I was in .NET land? Anaconda has made working with python a pleasure.

I'm really eager to start the second course in the series (regarding Regression), but first I want to spend some time to familiarize myself with both Pinswandas and scikit-learn. I'll then write a post to "translate" all of the homework assignments done in Graphlab-Create to use these two open-source libraries.