Title is a pretty good TLDR.

I would like to get a PhD sometime in my life.

I have a stable job where I can work from home frequently. Unless something catastrophic happens, I should be receiving my BA in mathematics this summer.

I would like to continue my education. I know of many data science and machine learning online master's degree programs that I would like to apply for.

I would love to be able to study some particular aspect of machine learning and become an actual expert in it one day. Automated machine learning comes to mind, but many other aspects fascinate me.

I'm perfectly fine being an older student at any college but there are some issues that are getting in the way of my long-term goal.

Right now I'm the main source of income for myself and my girlfriend. So I will have to wait until I am more financially stable before I could apply for a PhD program. From what I understand, most programs do not offer much income. It's going to take a couple of years of planning and sacrifices to be able to financially handle a lower income.

In the meantime, I feel that getting an online master's degree would be helpful. The problem is that almost all online degrees DO NOT OFFER A THESIS OPTION. This concerns me, since that would be the most enjoyable aspect of grad school. However, if that is my only option, I am okay with that.

Would it make sense to go to an online grad school and, when I'm financially stable, cross my fingers that a PhD program will accept a student who has only attended online classes for the past 3-5 years? From what I understand, most PhD programs want to see some research experience. Or at least have some professors that you can connect with. Going to school online prevents me from making a connection with professors, so I have doubts that I could get strong recommendation letters that way.


  • What can I do with the next 3-5 years of my life to help my goals?

  • How can I connect with professors who I only interact with online?

  • Is it even possible to do quality research while working full time?

  • Should I wait until I'm financially stable and then apply to a physical campus for a master's program?

  • Are online degrees even worth it, if my ultimate goal is a PhD?

  • (current phd student here) What is your goal after getting PhD? Do you want to be employed in ML field? Why your desire to "become an actual expert" can't be satisfied by self-studying and taking online classes or local college classes? Jan 14 '19 at 18:54
  • 2
    I won't go too in-depth into this, but the honest truth is that many universities are hopping on the "data science" or "machine learning" or "business analytics" online bandwagon. Many of these programs are extremely watered down and will seem as such to someone who has a bachelors in mathematics. So, in your choosing of a program, I would recommend that you be wise about which programs you look at. Many of them don't assume much of a prerequisite background. (For instance, I know of some programs that don't even require calculus.) Jan 14 '19 at 19:06
  • Thank you @Clarinetist This has been my observation as well, which is why I thought I'd ask to confirm how I felt. Jan 14 '19 at 19:14
  • @aaaaaa Doing work on your own is good but let's be honest here. A PhD is a serious commitment only for someone who seriously wants to create new knowledge for humanity. As much as I can picture myself somewhat happy with just a masters, I have a strong suspicion it might not be enough. My goal with this post is to find out what is the best path, not necessarily to decide prematurely. Jan 14 '19 at 19:15
  • 2
    @FarleyKnight An online masters program is almost certainly meant to be a professional masters, to teach students to apply some techniques from machine learning to problems in their own field. Such a program is unlikely to be designed for someone intending to do research in the field of machine learning, similar to how a course teaching how to use MS Excel is not going to be designed for computer science students that want to create a new database architecture.
    – Bryan Krause
    Jan 14 '19 at 20:13

Rather than give a long answer to a long set of questions, let me propose a strategy for getting all your answers fairly quickly. Make an appointment to visit a quality university and a professor in machine learning. It might be a place you'd want to go, but it need not be. Ask him/her these questions and try to work out a plan that might lead you to success.

I'm going to guess that online-only education will be less valued than in-person education, but by how much, I can't say.

Yes, you can do quality research part time, but it takes a lot longer than when done full time - especially if you don't have collaborative relationships to aid you.

The other questions (and even these) should be explored face to face with someone who can answer follow up questions and help you devise a plan.

Note, too, that 37 is still pretty young. Ignore the age factor.

  • Thanks for your response! I actually have talked with professors before about this. Unfortunately, they mostly have given me the same answers you have. Jan 14 '19 at 15:07

I have been fortunate to be in a top online M.S. Statistics program which is synced with the on-site M.S. Statistics and Ph.D. Statistics students, am almost finished with the program, and I've considered a Ph.D. in the past, so I feel like I can offer some insight.

TL;DR I should have done the Ph.D. coming straight out of my undergraduate. Doing it after an online M.S. can be done, but you will have a lot working against you.

For me, the online education aspect isn't an issue. My degree will be identical to what I would have gotten if I were on-site.

I am expecting no letters of recommendation from my M.S. program. Although I've been pleased with how the online content has been delivered in coursework, I haven't had opportunities to interact with professors. The main benefit of the M.S. content has been to be able to implement what I learn in my full-time positions in data analysis, and to be able to learn some advanced material on my own. This, of course, is not the point of a Ph.D.

Another difficult part of online learning, in particular while working full-time, is that employers will be mixed toward supporting you toward your degree, whether that be letting you take vacation or offering tuition reimbursement. I am aware of some anecdotes of hiring managers that will immediately reject candidates who are currently pursuing graduate degrees, because they feel that candidates will leave as soon as they are finished with them, and that such candidates do not apply to jobs to benefit companies, but rather, to only benefit the candidate's self.

I also don't interact with others in the class, other than a periodic e-mail to a few people whom I know. The M.S. program has taught me to become very independent and how to learn complicated material on my own, but one of the parts that I miss about my experience with the on-site degree that I have is the interaction aspect: interacting with other students, as well as professors, and building those relationships. This matters in more ways than one might imagine, not only for letters of recommendation, but also for attending seminars (I don't live anywhere near the campus which houses the program I'm in), and also for finding new (even paid (!)) opportunities. As one example, one of my peers did a paid government project due to a connection she had with a professor; that is an opportunity that I would not have access to.

  • I think you bring up some important points. Fortunately my current job is working for an online education company, and they do offer some data science masters programs online. I believe I will be eligible to pursue one of these programs completely free. At the same time, some of these programs, as @Clarinetist mentioned, are not as rigorous as I'd like. If it were possible for me to jump into a PhD program immediately after my BA, I would consider that track, but financially I won't be able to do that for a while. I think the networking aspect is something I will have to take care of on my own Jan 14 '19 at 19:54

I was chatting recently to a new PhD student in my university, who had recently taken her Bachelor's degree. When I asked her why she was working on the specific topic of her research, she replied that that was what her supervisor had suggested.

Nothing wrong with that, I suggest, if you are in your early twenties. But if you are a lot older, you might be expected to have in mind some aspect of your subject that you passionately wish to know more about and to contribute original work towards.

You will also, having more life experience, know more about how to work independently, and how to plan work over longer than the length of a course. An online masters will not give you anything more in those respects.

Do you know now what area of ML you would like to research? If not, you need to find out by learning more about the subject, for example by studying an online masters. As the holder of no fewer than three masters degrees, however, I feel that earning them required me to learn too many things that were not interesting and too little about the things that I did find interesting.

So, in short, there may be better ways than an online masters to follow your interest in ML until you think the time is right to seek a place on a PhD programme.

Not the answer you're looking for? Browse other questions tagged or ask your own question.