Back to School


Then big thing you forget about math classes is the problems. Every section of the book has 100 or more problems in it, presumably so the teacher can select ones out of the list and give them to students to develop their skills and an “intuition” about them (a term I keep hearing but have yet to experience). They start pretty easy and get harder and harder. My teacher likes to start our assignments with a handful at the beginning and a lot in the higher numbers, and we almost always do the last problem in the chapter.

My particular class covers three sections per week and does about 40 problems per section, so I’m doing about 120 calculus problems per week. The homework is turned in every other week, so aside from doing them to learn the material, I also need to do them for points.

That’s in addition to two quizzes per week and a discussion forum where we have to post our own interpretation of some calculus concept and discuss what others have posted.

Of course, in order to do the problems, quizzes, and discussions (which they wisely hide other people’s answers until you post yours), I also need to watch the hour-long M-W-F lectures and usually supplement them with a little Khan Academy and 3Blue1Brown.

I love my iPad Pro for this class… I can just write all my homework on graph paper using the Apple Pencil, then export the PDF. I can write directly on the quiz PDFs. I don’t have to deal with pencil smudges all over my hands and scanning it all in.

One plus side about this class is that the teacher posted all the work for the entire semester except the midterm and final exam, so if I can hang on by the skin of my teeth during the week, I can catch up and work ahead on the weekends. Oh, and I get to use the extra-expensive textbook for the second semester of this class as well.

Luckily my Intro to Programming class is relatively easy so far, but that’s mostly because I spent two months taking Python classes through Coursera, so I know the basics pretty well, I already had a development environment set up and a small collection of things I’d written already to use as reference. Hopefully this class stays this way.

That instructor is not so kind, and only releases the work for each week on Sunday afternoons, which is then due the following Saturday.

My third class, Data Analysis with Python, starts tomorrow, so we’ll see how that goes. I told myself I could load up and always drop the class if it was too much.

All told, it was a fast start to the semester. If I can survive this pace, I’ll have all six of my prerequisites done in two semesters and I’ll also have a certificate in Python programming as well.

Just to update the university search:

I’ve been accepted to Johns Hopkins Masters Science in Data Science program. They are not kidding… multivariate calculus is a prerequisite for almost every class. It’s a great school, but the list of classes available doesn’t fit quite while I’m looking for. They seem to specialize in bio statistics, while I was looking more for business and AI. Since I would not be able to take multivariate calculus until next summer anyway, they are kind of a safe fallback (and will NOT look bad on a resume).

I’m still waiting to hear from Northwestern. Their decisions are due late October/early November. I would have to say they are my top choice, with a number of classes on both business and AI that I haven’t seen elsewhere. If I get in there, I would most likely do the master’s degree in Data Science followed by a post-graduate certificate in Artificial Intelligence. That way I can take more foundation courses and take all the AI classes I wanted (which I couldn’t figure out how to fit into the major otherwise). Interestingly, the courses for the post-grad certificate are about half as much taken that way, so in two more semesters, I can take four courses for the price of two. I figure the whole combination would take two calendar years.

If I don’t get into Northwestern, my plan is to finish out my prerequisites, take the GRE, then reapply to both Northwestern and Berkeley. I don’t have any classes in December, so that would be a great time to study for the GRE. Hopefully between the prerequisites and a decent score, it would make me a better candidate.

Harvard’s data science program is still an option, but I don’t like how long it will take because of their calendar layout and that there’s a risk of having to go to campus for three weeks for the precapstone planning. That would rule out the program since I really can’t take a month off and go hang out in Boston.

An interesting ringer is Penn State’s Master of Applied Statistics. Since so much of data science is applied statistics (arguably, data science is the combination of statistics and computer science). it could be an interesting alternative. They also offer a four-course certificate, which could be a little enhancer, too.

So, while I’m hammering away at the prerequisite work, I’m keeping an eye on where it’s all going!