I'm going to be applying to grad school (a combination of direct PhD and research thesis masters programs) this fall, and I cannot make up my mind between two different fields - machine learning (essentially artificial intelligence) and computer architecture (essentially processor design). They would fall under either the Computer Science or Electrical/Computer Engineering departments.

I obviously need reference letters to apply and I'm going to be asking professors I've done research with in the past. However, I'm worried that asking for reference letters for two very different fields reflects poorly on me; it makes it look like I'm not entirely dedicated towards either.

I'm afraid that the letters I get from the professors will be a lot weaker than if I only applied to one field, because in that case it would look like I'm extremely passionate about it, compared to now where it looks like I'm not entirely sure what I want to do.

I'm wondering if this is a valid concern, and if so, how I should best address it. I could obviously pick one field now, but I'd prefer to put it off till after I get offers, so I have more time to make up my mind.

Just for context, out of the three professors I'm going to ask, one of the professors worked with me on a machine learning project, one on a computer architecture project and the last one on a signal processing project (not directly related to machine learning or computer architecture).

  • 1
    Have you considered applying to schools that handle both machine learning and computer architecture in the same Computer Science department? Commented Apr 22, 2018 at 5:58
  • 1
    Does it look bad to whom? And as opposed to lying about your interests?
    – JeffE
    Commented May 10, 2018 at 20:23

5 Answers 5


Based on the title of your question, I thought you were going to apply for two really different fields like French Poetry and Algebraic Topology! So, while I appreciate the difference between software design and hardware design, I certainly would not characterise machine learning and computer architecture as "very different fields". Both involve understanding and using computers, and I'm sure there are many people with an interest in both fields. If anything, your interest in both fields reflects a holistic interest in the field of computer science and would probably be interpreted as a positive.

Regardless of how similar or different these fields are perceived to be, I cannot imagine that it would reflect badly on you to have an interest in both, and to be willing to pursue either field at graduate level. While academia tends to push professors to become hyper-specialised, it is not expected that we will have worn you down to this point before even entering grad school!


I did exactly this. When I applied to graduate school, i applied to multiple schools for both computer science and computational biology. My reasoning was that, while I loved CS and was interested in theory, my background was more oriented in biology, specifically neuroscience. Computational biology to some degree was the more logical step.

I got into good programs in both fields, so I don’t think the seemingly disparate fields shaped my recommendations in any negative way. Why? While it certainly helps to have a clear idea of what you want to do with your degree, I don’t think it’s always required. Sure, if you’re planning on being funded by a specific lab for doing specific research, you should definitely have a developed interest. But for a self-funded MS or a PhD with a more flexible funding structure, there will be a degree of exploration — and it’s part of the journey.

I ended up pursuing an M.S. in computational biology and doing research on the analytics pipelines used for genomic variant detection in cancer. Now I’m in a PhD program pursuing research in a very different direction, more akin to theoretical computer science and the impacts of computational complexity on certain problems in biology.

Ideally, your recommenders should not just know you in terms of your academic performance. They should know you personally and be able to speak to who you are as an individual as well. If this is the case, regardless of what programs you apply for, they will be strong advocates.


No not necessarily. Data, information and knowledge is not divided up into mutually exclusive "boxes."

Some universities now offer double master's master's degree programs. I seriously considered doing this at the University of Oklahoma. I already have B.S. in with a major in geography and concentration in physical, not cultural geography, and M.S. in geology from OU. I also have a B.A. with a major in American History and honors in history of science, and a B.S. in earth sciences/geology from Tulane University. I did not double major.

I would like to earn a master of library and information science from OU. I want to create or manage digital databases. I have no desire to mage other people's lives in any way. I've walked away from an MBA and MIS degree programs because they were the wrong degree programs for me. I did research and found that first I had to take 18 credit hours of courses for one master's degree before I could start taking courses for the other master's degree, then I found out I could take 9 credit hours of courses outside library science courses in the MLIS degree program. That meant getting an M.S. in computer science through the engineering college was probably not necessary for me. Now my only problem is my age (65) and a lack of physical stamina, which I can do something about. I am not sure, at 65 years old and counting, if I want to invest another 3 years of time into getting a second master's degree and be unemployable because of my age. Age discrimination does exist for common sense reasons. I might end up in assisted living before I could graduate. That's why I gave up on going law school. No lawyer or law firm would hire someone as old as I am as law clerk or intern unless it was a "pity" hire.

Maybe you can find someone who is willing to advise you on an interdisciplinary masters and Ph.D. The one thing I would warn you about is getting "caught" between departments or schools and department like I did. All graduate students are pawns in the multilevel mental chess game that is college and university politics. My thesis adviser for the M.S was facing retirement in 6 years. I was his last graduate student. I specialized in geological remote sensing and GIS. I created a digital database as my master's thesis. The one thing my adviser asked me when I first met him was did I have particular study area in mind. My answer was "no". He is an expert in the petrology of the Wichita Mountains. He was VERY understanding that the amount and type of field work I could do was somewhat limited because I have been type 1 diabetic since I was 2 years old.

Everyone "customizes" their master's and Ph.D. degree programs to a certain extent, but the real question is usually whose research are are you doing: your thesis dissertation chairperson's research, or your research; the type of research you want to do.

  • 2
    Hi Carol. Welcome to SE.academia; I hope you enjoy it here. I noticed you are using a lot of all-caps in your posts for emphasis. In case it is useful to you, another way you can add emphasis is to use italics (put single asterisks around text) or by using bold text (put double asterisks around text). See here for more details.
    – Ben
    Commented Apr 22, 2018 at 11:49

As others have pointed out, the fields that you've mentioned are in fact quite similar, and it is completely acceptable to apply for different fields. I applied to 4 different departments (actually even two different departments in the same grad school in some cases), as different as Biology and Informatics! I think it all comes down to how you handle your applications and recommendation letters.

If you're applying to grad schools in the US, this might be easier, as you can even address the fact that you're interested in both ML and Computer Architecture in your SOP, if the department has faculty with research in both of these fields (not necessarily the same professor; I mean it has professors working on ML, and professors working on Architecture). But you should handle this well. It is ok (at least in the US) not to be completely sure about what field you're going to end up doing your thesis on, since in Academia, interests are quite fluid and in fact the committee always knows that your interest might completely change and end up not working on any of these two as it is not uncommon at all (but that does not mean that what your write in your SOP does not matter, it just means they may not be as strict as you think about you being dedicated to your current field of interest). I can't tell you what is a good way to address both of your interests in your SOP, but I'm sure your professors could be a great help if you can go and ask them (and if not, you can ask this as a separate question and ask for advice from the community here in StackExchange).

If you are not applying to US schools, in many other countries (such as most EU countries), they expect you to have a relatively clear idea of what you are going to do before you start your Ph.D. In those cases, you may want to separate your applications into two disjoint groups and even if you want to mention both of your interests, you should use a language that puts more weight on one (I'd recommend you ask for a others' opinion about this case, as I know very little about it). You can also do this for the US schools if you are applying to a certain school and you know which field you'll be pursuing in that school.

The reason you may want to mention both of your interests even if one is not directly relevant to the program you're applying to is just for your SOP to match the recommendation letters you get. If you can't get 6 letters (to divide them into two groups, one for ML and one for Computer Architecture), then you should handle the language you use in your SOP carefully so your SOP matches both what your professors write in their LOR and your the field of research of the faculty at the department you're applying to and in particular your faculty of interest.

In any case, I think it is important that you show a draft of your SOP to the professors who are writing your LOR before they write it. You should also explain your interests to them (I think they should be fine with the fact that you are interested in two different fields, but I think it's better if you get the opinion of a professor here about that). That enables them to write a letter that leaves room for some flexibility for you in writing your SOP.

PS: I applied to departments in four different fields, and I got offers from 3, but I had only one offer from two of them, and all my other offers and even interview invitations were from the other one. I think that was because my SOPs (and probably also the LORs) ended up having a biased language in favor of the field I studied my undergrad in. Even though that field was (and is) also my preference for grad school, but I did not intend to write a biased SOP. So you would want to be careful in writing your SOPs and perhaps ask someone else to read your SOPs and give their opinion on whether you sound the way you intend to sound or not.


Are you kidding me? These fields are closely related, so much so that you have a bunch of machine learning talks in any hardware conference, and in some of them (like the GPU Technology Conference) it's, like, a quarter or a third of the whole conference!

I thought you were going to say "Computer Architecture" and "Medieval Heraldic Artwork" or something.

Even if the fields were further apart - the main points to be consider would be:

  • Can you find an appropriate research subject?
  • Can you find a good academic supervisor (and get him/her to take you on)?
  • Can you find funding?
  • Will you be able to leverage your results and/or your efforts later on, once you're done?

If the answers are mostly yes, then - go for it.

You must log in to answer this question.

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