My Machine Learning Mountain — Reaching Base Camp
A 30 Day Writing Challenge
It’s Day 7 of my 30 Day Writing Challenge. If you’ve missed the previous posts, start here to follow my journey as I learn new techniques in preparation for my job in machine learning.
I’ve learnt a huge amount this week, and written a summary post every day. Today’s post isn’t going to be very long — my brain has been working overtime. I just want to take a moment to reflect on the week.
Machine learning feels like a huge mountain to climb. It looks like you need a ton of specialist equipment and piles and piles of data to do anything useful. Even a cursory glance at what the big players (e.g. Google, Amazon, Facebook, etc.) are doing will leave your head spinning. The truth is that you don’t need a deep understanding of mathematics or need to be that strong as a programmer.
As the cliché says, you climb any mountain one step at a time. There are lots of simple techniques you can use, which combine to be much bigger than the sum of their parts. You can start drawing insights from whatever data you have available, if you can figure out what it might be hiding.
I started the week with only a basic understanding of how artificial intelligence works. The kind of understanding you’d get by being interested and browsing Hacker News once a day.
Just 7 days later and I feel like I know enough to contribute to product development conversations.
- I know some techniques for preparing data ready for use in machine learning. The real world is messy, so it’s worth spending time trying to normalise data before it’s used.
- I know some techniques for supervised learning like Bayesian Classification if we know what outcomes we’re expecting.
- I know some techniques for unsupervised learning like k-means Clustering if we don’t know what the outcomes look like.
- I’ve realised the danger of trying to apply machine learning to every problem I encounter. Often it pays to really understand a problem before attempting to solve it.
- I’ve put a plan on paper for how I’ll build a team, focussing them around a common goal while demonstrating added value.
Next week I’m going to look at neural networks and the kind of problems that they suit. They’re much more of a black box, so the posts will be more ambitious in length and content as I try to explain what’s going on.
I’m going to have a day off tomorrow — I’ve earned it 🎉.
This is a post in my 30 Day Writing Challenge. I’m a software engineer, trying to understand machine learning. I haven’t got a PhD, so I’ll be explaining things with simple language and lots of examples.
Follow me on Twitter to see my latest posts. If you liked this article please click the heart button below to share — it will help other people see it.
