My case is the following, I have already got a MSc in Computer Science, but my field of specialization was Theoretical Computer Science. Actually I am really interested in the field of Machine Learning, but unfortunately I did not follow any course about that topic in my previous master's degree.

At first sight, I was thinking to follow a PhD that has some relation to Machine Learning; so in that case I should really follow some Machine Learning courses by my own. I have just put this thought for a while, mainly because I have not been so lucky in my last PhD applications.

In any case, I was wondering if I should follow another Master's degree in Machine Learning; and maybe because I have already got my other MSc, the time for completing this new MSc would be only one year instead of the two years that is the common duration of these studies.

Also I was thinking that maybe I could take some single courses, for example some online courses, even though I am not really a fan of MOOC courses.

What it would be the best course of action?

  • Follow another MSc degree?
  • Follow some online courses in this topic? if that is so, which university offers them that the certification is recognized?

Updated info:

About my goals are the following:

  • Need of learn machine learning topics for academic purposes: teaching undergraduates and independent research
  • Some good PhD positions in which I am interested require that knowledge

Any advice would be great.


  • 2
    What's your end goal? I assume you're not aiming for degreeCount = degreeCount + 1. As stated, it sounds like you're simply considering taking courses for the sake of learning new material. There are a lot of ways to learn new fields without the expense of graduate school (see my answer to here).
    – eykanal
    May 1, 2014 at 16:47
  • @eykanal it appears your link has been removed, perhaps due to over moderation. By chance does it exist elsewhere? May 12, 2017 at 15:26
  • 1
    @Lee - see my answer below, I edited in my (other) answer there.
    – eykanal
    May 12, 2017 at 16:20
  • Oh really where is that? Are you referring to the huge section of text below, clearly labeled with your name and picture, and positioned no more than 1 inch immediately below where I commented? Thank you @eykanal, I think it see it now :). May 12, 2017 at 16:28

3 Answers 3


Given both of those stated goals, I would not pursue additional coursework to learn the material. Your current masters degree is likely good enough for either teaching or getting into a machine learning PhD program, and no degree is necessary for independent research. I would recommend that you use other resources to learn the material.

EDIT: Given that the question in the above link was deleted, I'm reproducing it here:

There are many ways to learn new topics. Taking a university-level course is only one option. Other alternatives are:

  • read textbooks
  • online video tutorials (Khan academy, youtube)
  • online pre-recorded lectures (e.g., iTunes U, but many, many others exist)
  • talking to peers
  • reading publications (for specialized topics)

Taking a non-required university course is usually not the most optimal approach, for a number of reasons:

  • You will spend a lot of time learning material not relevant to your problem, simply because it's part of the syllabus and the teacher found it interesting
  • The requirement to pass will require you to spend likely (disproportionately too much) time doing homework, studying for tests, etc.
  • It is very expensive

This isn't to say that self-study doesn't require homework, or that learning topics not directly related to your research is a bad thing. I'm simply suggesting that by pursuing more self-directed learning opportunities you will be able to make better use of your time.


Getting the job really comes down to "what you can do" (and sometimes "who you know"..). Degrees are nice (and necessary), but you're current degree combined with working knowledge of machine learning and a couple of certificates from MOOCs could really go a long way. You would save time and money in my opinion.

  • I think, considering the way academia seems to work, that certificates of MOOCs would have practically 0 value. Especially when applying for a PhD. It simply does not guarantee any level of familiarity or proficiency with the material. MOOC certicates are nice on a resume for a college/university graduate to show they want to keep learning new material, but other than that...
    – Mythio
    May 2, 2014 at 14:27
  • 1
    So are you saying you need to be an expert in machine learning before applying to a program in Machine Learning? MOOCs have tons of value. Anything is better than nothing. I know of professors going up for tenure that have listed MOOC certificates, and they said the MOOCs wen't over well with the board.
    – user14808
    May 3, 2014 at 16:27
  • "So are you saying you need to be an expert in machine learning before applying to a program in Machine Learning?" No, i'm not saying that at all... I'm saying for a PhD application it is of very limited value because it doesn't have any guarantees. If there is someone with an MOOC certificate in machine learning and someone with an MSc. in machine learning, who do you think will get the position? (disregarding all other aspects for a second)
    – Mythio
    May 15, 2014 at 13:13
  • Nobody is comparing a MSc to a MOOC. Who do you think would get the position? Someone with a MS or someone with a MSc and a MOOC? Disregarding all other attributes, of course.
    – user14808
    May 16, 2014 at 23:50
  • 1
    Person with a better knowledge that can demonstrate it well in interview and demonstration processes will get the job either if it has MOOC's or Master's. But getting to the actual interview stage will be of course easier with MSc or good examples of your machine learning projects Mar 29, 2016 at 10:12

Some universities like ETH in Zurich (Switzerland) offer the "Master of Advanced Studies" degree for the undergraduates that have finished a given minimal number of additional courses with success, only one or two of them mandatory (assigned by a mentor). This would give the wanted additional expertise over shorter time.

Taking the whole master degree second time is a huge effort that may not be worth it.

You must log in to answer this question.

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