This article was originally adapted from a podcast, which you can check out here.
At the beginning of the new year, in Episode #538 I introduced the practice of habit tracking and provided you with a template habit-tracking spreadsheet. Then, we had a series of Five-Minute Fridays that revolved around daily habits I espouse, and that theme continues today. The habits we covered in January and February were related to my morning routine.
Starting last week, we began coverage of habits on intellectual stimulation and productivity. Specifically, last week’s habit was “reading two pages”. This week, we’re moving onward with doing a daily technical exercise; in my case, this is either a mathematics, computer science, or programming exercise.
The reason why I have this daily-technical-exercise habit is that data science is both a limitlessly broad field as well as an ever-evolving field. If we keep learning on a regular basis, we can expand our capabilities and open doors to new professional opportunities. This is one of the driving ideas behind the #66daysofdata hashtag, which — if you haven’t heard of it before — is detailed in episode #555 with Ken Jee, who originated the now-ubiquitous hashtag.
As a bonus, hopefully you picked a career in data science because you find learning to be fun — assuming so, then aiming to hone a new technical skill every day should be intrinsically rewarding for you too. This intrinsic reward is a shared characteristic of this week’s technical-exercise habit and last week’s reading habit.
Beyond aiming to learn something technical every day, I’m not sure the specific subject area matters as much. The key, again, is simply to keep learning. That said, you might be curious why I choose to focus on math, computer science, and programming as the subject areas in my daily-technical-exercise habit. Well, these three subject areas are:
As detailed in Episode #556, the central underlying foundations to being an outstanding data scientist or machine learning practitioner
Often challenging, which makes pursuing them rewarding
Interwoven with each other, and for me discovering these interdisciplinary connections is especially exhilarating
Effectively infinite in scope, meaning in all the years I’m alive I’ll never run out of a wide variety of exciting new math, CS, or programming topics to dig into
So hopefully I’ve convinced you that this could be a great habit to adopt. If you happen to be looking for an interactive online resource to get you going, in no particular order my top three recommendations are:
Khan Academy for math concepts in general. They also have programming content, but it’s mostly focused on HTML, CSS, and Javascript so not necessarily the most directly relevant to data science (though if you’re keen on data visualization or user interaction, these are some of the top programming languages to know).
For interactive data science-specific education, I’m a big fan of DataQuest, so you can check that out.
If you’re keen to learn linear algebra or calculus for machine learning specifically, then you can undertake my free, hands-on intros to those two subject areas on YouTube (linear algebra here; calculus here).
And then if you’re looking for book recommendations, if you’re a regular listener then you’re already aware that we end almost every guest episode by asking the guest for a book recommendation. For your convenience, we’ve collated these recommendations for you in the SuperDataScience Podcast Virtual Library, which you can access for free at superdatascience.com/books. That’s superdatascience.com/books.
In terms of specific technical books that I recommend:
For general data-science modeling approaches check out the classic Introduction to Statistical Learning by James, Witten, Hastie, and Tibshirani
For digging into the relevant math of computer science specifically, I recently discovered Graham, Knuth, and Patashnik’s proper university textbook Concrete Mathematics to be delightful.
And, finally, if you’re interested in deep learning in particular, then of course I’ve gotta plug my book Deep Learning Illustrated.
Like the other habits I’ve already covered in my Five-Minute Friday episodes on my daily habits, I choose to log my “math, CS, or programming” habit as a binary habit — either I work through at least one exercise on a given day or I didn’t — so using the habit-tracking template I introduced Episode #538, I set the min column for this “math, CS, or programming” row of the spreadsheet to 0 and the max column to 1.
All right, that’s it for today. I hope you found this episode to be practical and I’m looking forward to catching you on another round of SuperDataScience very soon.