Last week I had the pleasure of attending a workshop on machine learning put on by Divergence Academy. They offer a lot of courses related to data science, including the workshop I took. They also offer more intensive courses on things like applied analytics and how to use specific tools like Hadoop, Spark, and various Python libraries.
Of course, big data is kind of sexy right now, so it would make sense that lots of places are offering training opportunities like this one. What did I enjoy so much about the training I got? The difference: the right mix of theory and hands-on experience.
Our instructor started out with an overview of the theory, but then we got a chance to apply the principles we were learning. We were shown what to install on our computers, and then we got a chance to play with some sample code.
Sometimes hands-on training can get very cookbook-style: “type x command and y will happen.” That approach can be satisfying, especially at first, but it has its limits. You don’t always get a sense of the overarching thing you’re doing. Fortunately, that wasn’t the case with this workshop. The instructor made it clear how the steps we were applying fit into a larger framework. It’s the difference between following a recipe to the letter and actually learning how to apply various techniques in the kitchen.
I want to unpack this idea a bit. It’s common, even popular in some circles to criticize learning that is too theoretical. “Theory is useless without practical application,” some say. Even if that’s true (which I don’t necessarily grant), let’s take the flip side of that situation. I’ve seen and experienced situations where people are walked through a process step by step. For instance, there are a lot of tutorials that can walk you through a particular program. If you follow the tutorial correctly, you’ll succeed at the task at hand, but unless you have some idea of the larger context, you won’t be a better programmer. Theory gives you some of that larger context. True, it can be dry sometimes, but a lot of innovations started out as theoretical ideas.
Now, since this was a one-day workshop, we were limited in how much theory and practical application we could get. One student mentioned they had hoped to learn more about deep learning, but for a one-day workshop, I think the content mix was about right. I pointed out to this person that universities have semester-long courses on machine learning, so you have to be realistic about what you can expect to cover in a day.
So if you’re in the Dallas/Fort Worth area and looking to brush up your data science skills, check out Divergence Academy. There are lot of fly-by-night training companies out there, but these people are the real deal.
Longtime readers of this blog (hi Dad!) might wonder if this signals a shift in direction for me: am I gearing up for a career in data science? Probably not, at least not as a primary motivation. I’m interested in how I can apply some of the concepts to data science in other fields, though. Of course, data plays a big role in UX. Also, journalists can use data to support their stories.
In fact, that was one of the things we talked about at the workshop. It’s all well and good to have the numbers, but if you can’t tell a story with them, maybe those numbers aren’t so useful. If that isn’t connecting technology and humanity, I don’t know what is.
What (good or bad) training experiences have you had? Share them in the comments!