Eric J Ma's Website

On Learning Math

written by Eric J. Ma on 2017-01-05 | tags: data science education graduate school


Though I will admit to being somewhat algebra-blind (more on that later), I wasn't necessarily bad at math concepts. I did have one big problem with the way I was learning math, though - it always seemed to be more theoretical and less applied, as if solving puzzles for solving puzzles' sake was the best way to approach learning math.

I'll grant that, yes, many mathematicians and statisticians I know, are one of those types where doing and learning math doesn't have to be tied to some applied goal, and for whom math is viewed as intrinsically fun. But... I'm not one of those types, and it was for this reason I nearly abandoned quantitative thinking in my undergrad. Math was boring after Science One, because I wasn't shown how it was applicable to real-world problems in upper-year courses. Yet, after this PhD, I'll have essentially developed an area of expertise in some hybrid of statistical evolutionary biology and deep learning for biochemistry. Makes me wonder whether the current mode of puzzle-/theory-driven math (and perhaps even CS) is actually driving off people who might not be interested in the intrinsic fun of math, and are more interested in the applied fun?

Science One math was taught by Leah Keshet and Mark MacLean. The part I remember most vividly was learning about differential equations as applied to ecological problems and biochemical reaction kinetics. Along the way, we had to learn differentiation and integration... not for differentiation and integration's sake, but for figuring out useful properties of these two wildly different systems. Now that was amazing math.

What do I mean by algebra-blind? I basically mean that I tend to confuse algebraic symbols, and can't hold them in my head without having some lookup-table/glossary for what they mean. The lookup-table of algebraic symbols to their meaning was a hack that I never mastered until I finished formal math education; yet it was extremely helpful to have when I was discussing math with my collaborators up at Harvard. Most of my score deductions in math tests came from confusing algebraic symbols with one another...


I send out a newsletter with tips and tools for data scientists. Come check it out at Substack.

If you would like to sponsor the coffee that goes into making my posts, please consider GitHub Sponsors!

Finally, I do free 30-minute GenAI strategy calls for organizations who are seeking guidance on how to best leverage this technology. Consider booking a call on Calendly if you're interested!