8

Say I wish to pursue a phd in some CS related field. How knowledgable do I have to be in the field I am pursuing? I'm asking because some things happened recently that led me to doubt myself:

I have been doing some machine learning research for the past year and I'm working towards a first author conference submission (as an undergrad). As I was drafting my paper, my professor asked some questions about the project I'd done - such as why I chose certain methods over others, or why I choose a particular parameter for my experiments and more. And I realised that I couldn't answer them. See, in my project, I didn't create anything new. I simply took existing methods and applied them to a relatively unexplored research area, which means that I was just following what other people did. To take something and apply it somewhere else doesn't really require that much knowledge in my opinion. I've been reading papers, trying to explain why some methods outperform others and often times I do not fully understand them. The high level logic maybe I understand, but certainly not the in-depth algorithms or implementations.

I feel as if I know nothing about this field (and I probably really don't know much). Yet, my professor offered me a phd position. I feel like he was placing too much confidence in me. Have I managed to fluff my way through this project, or is this something that happens to other people too? Especially those who completed their bachelors and went for a direct phd.

  • 2
    I barely know anything about applied probability and optimization but that's where I am. You'll learn about the subject area as you move along. – Sean Roberson Aug 31 '17 at 15:38
  • 5
    With a few rare exceptions, first year phd students are mostly useless :) – Fábio Dias Aug 31 '17 at 15:47
  • 1
    Talk to your potential advisor. If possible, organize a meeting and have a chat. Granted, the formality of the occasion will depend on social norms in your country, but it should be possible to have an open conversation about any apprehensions you have about the position. If nothing else, the conversation will give you more information about what to expect during your PhD, which is not an outrageous request. – Will R Aug 31 '17 at 17:22
  • 2
    "I simply took existing methods and applied them to a relatively unexplored research area". You've just described how most research fields are born. – user63725 Aug 31 '17 at 17:24
  • 1
    Obligatory imposter syndrome link. – Mad Jack Aug 31 '17 at 17:46
4

It is very appreciated for your honesty. In my knowledge, I have seen many students they join PhD courses having very little knowledge in their field. Also, many join in new fields without having any proper/concrete ideas. However, their interest to learn, hard work and self confidence all together help them to achieve a PhD and a great career ahead. It is also worth to mention that each PhD student must face lots of ups and downs. However, their self confidence and interest to do something new helps to overcome all the down situations.

So first of all, you should never loose your self confidence at the very beginning. It is very well said that "it is never too late to start". You will get enough time in PhD to learn many new things including your objectives. Moreover, it seems you have got a very good and inspirational professor. So there is no doubt whatever knowledge you have now, it is very much enough to pursue a PhD in your area of interest. Good Luck!!

2

It appears you may be suffering from 'imposter syndrome' - welcome to academia! This particular ailment is widespread, I'm afraid.

Especially as an undergrad, no one expects you to be an expert. That's why you go to graduate school. The fact that your advisor asked those questions suggests he/she thinks highly of you, and suggests he/she may be a good advisor.

Now, and in the future, use these situations to identify holes in your knowledge, then fill them.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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