15

I'm a beginning doctoral student in the field of Machine Learning and I've been given several contradicting opinions on whether a PhD is worth it or not. I don't want to spend the rest of my life in front of a desk doing research, but I would like to deepen my understanding on my field. I am interested for a career on industry. Some people say I'm wasting my time for doing the PhD and some people say it is worth it. My plan at the moment is to finish PhD as fast as possible and then continue my career in industry when I'm still young. Any opinions or recommendations on what I should do? :)

P.S.

I want to add to this that I have already 2,5 years of work experience on industry (IT) already. For example I designed and implemented an online language skill tester for my university and have worked as a software engineer. So I wouldn't be entirely newbie in programming after I finish my PhD...

My subject is about applying the techniques of Machine Learning on Big Data.

Thank you for any suggestions and guidelines :)

3

6 Answers 6

19

You haven't really defined what you mean by "...when considering your career" in the subject of this post. On the one hand, you say,

I would like to deepen my understanding on my field.

If your goal is to deepen your understanding, one way to do that is to get a PhD. You can also do this on your own, without getting a PhD, but you'll find a more structured path and (possibly) more resources if you go down the PhD route.

On the other hand, you say,

I am interested for a career on industry.

Are you interested in industry because of the money? If that is the case, you're most likely better off earning a full income and "moving up the ladder" during the time you would spend in graduate school, which means you should forego getting a PhD to concentrate on the money.

If you're interested in a career in industry because of reasons outside of financial considerations, then you have to think about what kinds of jobs you're looking to take in industry. If you're satisfied that your current skills and educational level will make you competitive for the jobs you want, you should probably reconsider the PhD, and to "deepen your understanding" in a different way. Then again, a certain percentage of people want to get a PhD for the challenge and for strictly personal reasons, removed from the job details themselves.

If the jobs you want to get are generally given to PhDs, then you have your answer already -- get the PhD to make yourself competitive for those positions.

I suggest writing down all of your long-term goals, and then weighing them against the time and cost (and opportunity costs) of getting your PhD. If after all that you still don't have a good answer, you might consider continuing with your program for another year and just postponing the decision until then. Eventually, either you will get your PhD, or you will be convinced that it is better off that you stop the program (but the decision may get harder the closer you are to finishing).

0
13

It is impossible to answer your question unless you tell in which country you intend to work in the industry.

In France, for example, a PhD is not very highly considered by industry, where the one important criteria is which engineering school you did. This is slowly evolving.

I heard that in Germany, a PhD is highly valued in industry, even if its content has nothing to do with the job.

I have little knowledge for other countries.

1
  • 4
    Re: Germany. It is. The idea is that a PhD holder has at least managed one large complex project till the very end: his/her own PhD project. Jul 21, 2018 at 23:03
6

I completed my PhD recently and I am currently working in the industry. Mine is a PhD in CS and my thesis was on development of newer techniques for real time prediction. Before PhD, I had 5 years of work experience in the analytics sector. Personally, I found the entire experience of the PhD to be intellectually satisfying and worth the effort. Yes, I had to forgo 4 years of income or earning opportunities and survived on an adequate stipend. But I believe PhD in a technical field like Machine Learning will always be in demand in the industry. Post PhD, I got the job as a data scientist with the largest online retailer. I think it varies from individual to individual and one should find their own way. In my country, PhD is not that highly valued in the industry. But I guess one should do a PhD in an area for which they have some passion so that they can develop the skill sets that will make them competitive in the job market.

0
5

This is not an 'official' answer, but one of a revelation of sorts I have had recently.

One thing that occurred to me as I complete my PhD (submit in the next fortnight), is that my PhD is a very practical and successful application of existing technology into very specific scientific functions that the technology was not designed to do. This, and the community benefit aspects have formed the basis of the papers that I have been able to get published. If you can make a focus on practical applications alongside the theoretical constructs - then the PhD may be of benefit of both you and a future employer.

However, this varies from discipline to discipline (as the answers in my other thread linked in the comments suggest).

You have got to look at what benefits are there for you in pursuing a PhD, both in the short and long term.

0
3

If you have an opportunity to finish your Phd then do so. Failing to continue with a Phd is something that is difficult to change your mind on later in life. In the big scheme of things what you see yourself doing just after graduating is so different from what you may find yourself doing in 10, 20, 30 or 40 years later.

You will never regret having a Phd. You may regret not having a Phd when you could have completed it!

1
  • 1
    Thank you Brian :) appreciate your help. In fact, I recently graduated as a fresh PhD :)
    – jjepsuomi
    Jul 20, 2018 at 16:21
1

In my opinion, the key defining feature of Ph.D. research is to understand a subject in its most fundamental aspects and push the boundary of collective human understanding and knowledge. There are some things that you only get to explore while going through a Ph.D. degree and some results you only get to publish in academia.

Consider what you have said in your post:

I don't want to spend the rest of my life in front of a desk doing research, but I would like to deepen my understanding on my field.

The question is what do you mean by deepening your understanding of your field. If by deepening your understanding, you mean, deepening your understanding of the technologies that are used in machine learning, then industry could be a good option. But if you mean that deepening your understanding of how and why things work, then perhaps the industry is not a good place for that.

My plan at the moment is to finish Ph.D. as fast as possible and then continue my career in the industry when I'm still young.

I think you should examine whether you want to get a Ph.D. solely to satisfy some requirement on a job application, and perhaps to get a higher starting salary, or do you want to create something that lasts and make a name for yourself in the research community. The key question is: "will the impact of my thesis be felt long after I finish the Ph.D. degree"? And if your Ph.D. research cannot provide a satisfactory answer to this question, then it is better to go into industry.

I have observed that there seems to be a collective rush to finish Ph.D. in machine learning and computer science departments. In my opinion, this collective habit has greatly cheaped the value of their degrees, and this can be reflected in the quality of their thesis, which consists nothing more than a few equations, some hand-wavy explanation as to why their new equations are better, and some simulations with highly controlled experiments.

I have seen the statistics; nobody cares about their research at all (view count less than some double digit number, years after publishing thesis), because everybody else is also putting up a few equations and doing their own simulation and hoping to get into industry asap - the level of toxicity is high, nobody is building upon or examining each other's work. This kind of research is unsustainable in the long run.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .