So this is a question that do I need a degree in Machine Learning or not, which I get asked quite a lot, and the answer is it depends. It depends on you. It depends on your background, what you’ve done so far, what you want to do, what kind of career in machine learning you want and whether or not you want a PhD or masters.
Now, I’m not saying to get one, just that it might not be necessary. Let me explain. Let’s take a PhD. It’s a huge time commitment. It’s a big financial commitment. And unless you’re doing the right PhD at the right institution with the right supervisor, then it’s not going to guarantee you a good Machine Learning career.
Also, Read – Machine Learning Full Course for free.
Now, you might be lucky and it might help, but I think if you’re committing all that time and money, you need a better outcome, then it might help. Now, there is a little caveat to this. Some roles within the large companies will require a PhD.
Which is a Better Degree in Machine Learning?
Now, I know I just said that you don’t need a PhD to get into Machine Learning, but it depends on the sort of machine learning that you want to do. These roles that I’m talking about are scientific research roles where you’d be required to have a PhD in something like artificial intelligence or machine learning.
You’d be expected to have quite a few published papers to your name where you’d be the first author. And these are roles where the companies are looking to develop new algorithms or new ways of applying algorithms to certain situations. But they’re very specific. When most people talk about machine learning, they’re talking about applying existing methods to a bunch of data to gain insight from it and to be able to predict things from it. That’s the sort of machine learning tasks that I’m saying you don’t need a PhD for.
Do you want to do the other stuff then? Yes, absolutely. You will need a PhD. So what about a Masters, masters could be a good option, it can be a good way of getting a job in the sector, especially if your undergraduate degree didn’t have much romance in it, if it wasn’t a quantitative degree and if you haven’t done much coding.
But there are lots of machine learning masters degree programs springing up all over the place. So you want to make sure you choose the right one, choose one that isn’t too long. You don’t want to spend years doing your masters.
So if you can find a one year masters, that would be great. You don’t wanna spend too much money on it, but you also want to make sure that there is some guarantee of a job at the end of it. So before you apply for the Masters, make sure that you’ve contacted them to find out what sort of links they have with industry and whether there’s a placement and what sort of support they provide once you’ve done the Masters to get you a job, because that’s what you want it for.
What if You Don’t Have Any Degree in Machine Learning?
I said masters can be useful, but I don’t think it’s essential. So what other options are there open to you? Now, in the words of the renowned data scientist Jeremy Howard, there’s nothing that you could learn in a Masters that you wouldn’t be able to learn from books or online courses.
Now, of course, you won’t get the interaction with books and online courses that you’d get on a Masters and you might not get the choice of projects as well for your portfolio. But that doesn’t matter because there are so many organizations and companies that you could offer to do data analysis for. And they’d be very pleased to have you because many of the smaller organizations and companies can’t afford a data scientist.
And that way you can build up your portfolio, which will give you something to talk about when you’re interviewing for machine learning positions. And you’re going to need that as far as where to go to learn all of this stuff. Well, if you’re looking to learn it for free, you can do that. There are loads of resources online.
If you’re looking for something a little bit more curated, then I would recommend Coursera.
How To Get Your First Job in Machine Learning?
The next step is extremely important. Once you’re making progress, you’re working through the online courses, you’re reading your books, you’ve got a couple of projects under your belt, then you’re going to start applying for jobs. But you mustn’t apply for machine learning jobs at the very initial stage. You’re going to apply for data analyst jobs, choose companies where you’re not going to be the only person working with data within that company.
You need to choose companies where they have a team of people working on data so that you can be integrated into that team and you can work with people that are more knowledgeable than you and you can learn from them. And gradually you can work your way up from being a data analyst into more of a machine learning role. And I think that’s a really good way into the sector, to the industry.
But what if you don’t have any qualifications? What if you don’t have an undergraduate degree? If you’re in that position, there’s no doubt about it is going to be more difficult. What I would suggest is if you can get an undergraduate degree, if you can start an undergraduate degree in a subject that is related to machine learning, then do that.
If not, just learn as much as you possibly can about the subject and get as much experience as you can to as many projects as you can and try to get an entry-level job somewhere. A little time, effort and dedication, and I are sure that you will succeed.
I don’t think a PhD is necessary. A masters degree is useful, but not essential. What’s important is to have some projects that you’ve worked on, learn as much about the subject as possible, and then aim for an entry-level data analyst job.
Just one final thing to say now, and that is I don’t want you to think that I’m trying to put you off doing a PhD. There are loads and loads of great reasons for doing a PhD and a PhD can be a fantastic life experience. It can be a fantastic learning experience and it can help you with your future career.
All I’m saying is if your only motivation for doing a PhD is to get a job in machine learning, and I don’t think it’s worth the effort. But if you just want to do one anyway, that’s great. They’re fantastic things to do.