Nick Ang

Should I try learning machine learning (again)?

March 09, 2018

I believe some of the best moments in life happen serendipitously, when you just happen in the right place at the right time, with a positive disposition primed to experience it.

I’m now thinking if my trip to the local library this evening is such a moment. Of course, I’m jumping ahead. Let me explain.

Because my wife recently started doing video tutorials, she thought it’d be good to have nice looking nails. So she went to the nearby Nail Palace after our dinner, and I was (happily) left to my own devices.

I chose to go to the library, of course! Everything else in a shopping mall is just plain boring, perhaps with the exception of Popular, but only because they sell books…

Post-dinner library visit

Once I got past the front desks featuring the latest novels and travel guidebooks, I went straight for the Computer Science shelf.

I browsed for about 15 minutes and picked up 2 books - Introduction to Machine Learning in Python by Andreas Müller and Sarah Guido (website) and Eloquent JavaScript by Marijn Haverbeke (PDF). The whole time I reminded myself I wouldn’t possibly be able to read more than 2 books in any meaningful depth in 2 hours…

Seeing that I’m job searching now, I prioritised reading the machine learning book over the JavaScript one.

I’m already familiar with JavaScript and Python, but machine learning is something I’ve tried to learn before but failed because I found it too dry during my first attempt to learn it.

At the back of my mind, I also knew that machine learning engineers are currently in high demand while mostly remaining in short supply.

A good intro to ML book

introduction to machine learning with python book cover

Glossing over the content page, I picked chapter one and dove right in.

The next hour blitzed past in a flash! I was having a really good time learning supervised machine learning with clear examples and concise explanations in Python. Kudos to the authors for keeping the language concise and stripping the examples down to the essentials!

So now, as I sit in the pedicure chair next to my wife writing this, I’m wondering whether I should dive deeper into ML.

Am el. Few months ago these two syllables didn’t belong in front of the word “engineer” to, because, well, I hadn’t even heard of it before! But now, “ML engineer” has become a real thing, and the demand appears to be steadily rising.

Prestige and trends

I’m a big believer of not making decisions based solely on prestige or trends though, because they completely ignore one’s interests, aptitude, and life situation. So I’m not going to dive into ML just because it’s going to get me a job. It obviously has to be evaluated on a more robust criteria, which is not within the scope of this post.

Now, one last thing. I’ve said this before and I’ll stand by it again: I want to love machine learning and artificial intelligence technology as an engineer!

Conceptually, I find ML and other domains in AI fascinating. But when I say conceptually, I really mean high-level, probably slightly misinterpreted ideas of what they are and can do. That probably explains why I continue to revisit it time and again…

The conundrum is whether to put in time to find out if I’m right about my ideas of what ML and AI are about. Time invested in this is time not invested in getting better at building web applications, and it needs to be a strategic decision, I think.


Written by Nick - I don't know but I'll learn (twitter)

Nick Ang © 2020

twitter iconinstagram icon