If you found this page, it means two things:
- You are interested in data science, but aren’t sure where or how to begin.
- My SEO skills are improving.
I might be able to help with #1. I can tell you what I’ve done so far and of those, what worked. Maybe you’ll pick up a tip or two, and I will be able to do the post title justice.
#2… Let’s leave it at that.
You might be wondering what credentials I have to say I can help you out. I don’t have any. I’m a complete data science newb. But, why would I need credentials for wanting to help?
If you know less than I do, I can teach you.
If you know more than I do, you can teach me.
If you and I are on the same level, we can help each other learn.
That’s the idea behind this post.
I won’t go into defining what data science is (and its related terms, such as big data and data analytics). I will assume that your finding yourself here means you’ve already googled all that. What I will do is talk you through what I’ve tried so far that has worked for me, and maybe those would help you too.
If you’re just starting out with Data Science*, I would advise you to:
1. Get a proper overview.
The great thing about data science is that there’s a lot of data about it (current Google hits: 58.6 mil**). The not-so-great thing about data science is finding where to start. I mined the Internet for quite a bit before finding some great resources.
One of those great resources is Springboard’s “Getting Your First Data Science Job” (review to follow). I’d recommend it to anyone starting out because,
- It’s realistic. The book goes out of it’s way to make sure you have an idea of what being a data scientist in the industry is like.
- It’s organized. The content, and the flow of the content, are well thought-of. I especially love the little checklist in the end.
- It’s FREE.
2. Have a personal project.
It helps to have an idea why you want to learn data science. Think about a question you want answered that needs data science to solve it.
Maybe you want to get the best price for your next ebay auction. Maybe you want to find out what makes popular music tic. Maybe you run a blog and want to find out what topics appeal to your readers *hint*hint*.
Whatever it is, keep it in mind while studying. I found that it keeps me motivated as I can find an immediate use for whatever new thing I’m learning.
3. Study with structure.
Speaking of studying, if you find from the overview that you have a skill gap, you will need to study up. But don’t just study! Study. With. Structure. Have a study plan. Know what you need to study about, how you’ll study it, and what’s a good order to study them in. This increases the chance that you’ll progress, and you’ll have a means of measuring that progress.
For example, I’m currently taking Microsoft’s Professional Program in Data Science via edX. The course structure and timeline are both predictable so I can plan my schedule around them. I can focus on the present course, because I know that someone credible (Microsoft) has already planned what’s next for me. The program and its courses both have progress bars so I can keep myself in check.
I found all this much more effective than say, borrowing a book on R, then wishing it would magically translate to a skill (hint: it doesn’t).
P.S. You can follow along my progress with the Microsoft Program here.
4. Avoid forums.
Information-rich sites like Quora and Kaggle are great if you use them for their purpose. Have a question? Ask Quora. Want to practice your skills? Compete in Kaggle. Simply browsing? Stop. Its easy to get overwhelmed by the information available. It’s easy to doubt the structure you set up for yourself.
Doubt has a large place in data science (hello, statistics!), but you’re just getting started here. You don’t want doubt. You want, no, NEED a map. That’s what the overview in #1 is for, that’s what the structure in #3 is for, that’s what this post is for.
5. Show up.
Once you get started, it will be hard to maintain that inertia. Maybe real life keeps getting in the way, maybe you’re starting to get disinterested, or maybe you just want to give yourself a break.
Whatever it is, have the discipline to follow through. Show up and do the work. Sean Wes has worded this so well at his blog, so I recommend you go over there and inspire yourself.
Personally I’m doing about an hour each day, Monday to Thursday. Sometimes I do more, but I don’t ever do less. If I find that I can’t make it to one of those days, I try to off-set by doing extra hours another day. It’s hard, but it works. I know because I just got my first course certificate last week, and am about half-way to my second one.
That’s it. Nothing new or earth-shattering, just some solid advice. While I had data science in mind here, the overlying tips will apply to getting started at almost anything. I’ll round it out at five because, “5 Tips on Getting Started with Data Science” is perfect SEO material… right? And as with most advice, YMMV. Feel free to suggest some of yours as well.
*I wrote this with self-studying in mind, but I daresay it’s applicable to anyone.
**This amount of data is just begging for an analysis, in case you’re still looking for your own personal project. When did “data science” as a search term start peaking? What peaked around the same time–is there a correlation? What’s the demographic of the searchers? Is any demographic more statistically significant than the others? And if so, why? Based on these, can you predict what the next generation of data scientists is going to be like?