The 5 Must-Read Books for the Self-Taught Data Scientist

by | Sep 13, 2018 | Self-Teaching

Take your pick of articles related to learning about strategies used by the uber-wealthy and you’ll find a common theme: spending at least five hours per week on deliberate self-learning. Try these on for size:

The point is that this practice is deliberate, meaning that it’s done consciously and intentionally.

Over the past few years, there have been several books that have been insanely useful to me when self-teaching myself any topic (from fixing my dishwasher to learning to use scikit-learn). Here’s my takeaway of the five I’ve found most useful.

The First 20 Hours: How to Learn Anything…Fast! by Joshua KaufmanThe First 20 Hours: How to Learn Anything…Fast!

by Josh Kaufman

Kaufman makes the case that you don’t need to do something for 10,000 hours in order to become an expert. Instead, he suggests you can complete “rapid skill acquisition” in 20 hours through his 10-step method:

  1. Choose a lovable project.
  2. Focus your energy on one skill at a time.
  3. Define your target performance level.
  4. Deconstruct the skills into subsets.
  5. Obtain the critical tools.
  6. Eliminate barriers to practice.
  7. Make dedicate time for practice.
  8. Create fast feedback loops.
  9. Practice by the clock in short bursts (such as with the Pomodoro technique).
  10. Emphasis quantity and speed.

In my opinion, Kaufman’s is the best process for any self-teaching endeavor, and I encourage you to keep this list in mind as you read through the rest of the book recommendations, since all of these books can be used to enhance certain steps.

A Mind for Numbers by Barbara OakleyA Mind for Numbers: How to Excel at Math and Science (Even If You Flunked Algebra)

by Barbara Oakley

Think about it: Did anyone actually show you how to learn? For me and most people I ask, the question is a resounding no. In 2017, the third most popular class on Coursera was “Learning How to Learn: Powerful Mental Tools to Help You Master Tough Subjects.” A Mind for Numbers is a companion to that course and reading it really changed my approach to learning on my own, especially after reading her section on “chunking” (which she also discussed on Quora.)

Steal Like an Artist: 10 Things Nobody Told You About Being Creative by Austin KleonSteal Like an Artist: 10 Things Nobody Told You About Being Creative

by Austin Kleon

While I don’t condone stealing of any kind, you should definitely learn how to copy then put your own mark on your work. Within the context of this book, Kleon says that “copying is about reverse engineering. It’s like a mechanic taking apart a car to see how it works.” Building on this advice, self-teaching to get the tools and techniques for data science would mean finding a mentor or seeing how other folks are solving actual data problems. I tend to read this book multiple times a year since Kleon’s ideas are so relevant to the process of self-teaching — especially the concept of reverse engineering.

Deep Work: Rules for Focused Success in a Distracted World by Cal NewportDeep Work: Rules for Focused Success in a Distracted World

by Cal Newport

As I’m sure you can relate to, being able to focus your time on learning is easier said than done. Newport defines “deep work” as “the ability to focus without distraction on a cognitively demanding task.” His book outlines strategies for getting things done in a distracted world. Given that it’s easier to pop over to Amazon than to read the scikit-learn documentation (even though you need to if you really want to learn it), Deep Work is a must-read for anyone who wants to self-teach but is easily distracted.

When- The Scientific Secrets of Perfect Timing by Daniel H. PinkWhen: The Scientific Secrets of Perfect Timing

by Daniel H. Pink

While there are many valuable insights in this book, the one that stood out the most was that how we perform different tasks changes as we progress throughout the day. Pink states that our day occurs in three stages — peak, trough, and recovery — and recommends that we keep those stages in mind when scheduling activities. Peak time, for example, is best spent on analytical tasks, while busy work like clearing out your inbox should be saved for the trough. And creative work? Save that for the recovery stage.

And because everyone’s internal clocks differ, you also have to know whether you’re clock is set to normal, lark, or night owl. Night owls experience the progression of the stages in reverse (e.g., recovery, trough and peak), while most folks experience a peak at around 9 AM, followed by the through and recovery phases. Larks experience the sequence a few hours earlier than that.