Seeking advice for strengthening my PhD application

[heypeach]

After working as a software developer in Canada for a while, I've decided to pursue a Masters in CS in Canada and will start in Fall 2015.  My ultimate goal in terms of education is to obtain a PhD in Robotics & AI in top 10, or top 20 CS programs in the US. My experience is a bit different than most applicants. I just wanted to share my thoughts here and hopefully you guys could give some suggestions or point out anything that I may be mistaken.

 

My main reason for going back to school is that I enjoy R&D type of work. I knew I would one day head back to school even before I started working in the industry. My background was in embedded system engineering (I have a bachelor and a Master, both in ECE). I found myself more interested in computer science and have always been doing some side projects and taking online courses about computer science.  I consider myself decent in Math and programming. Bu I don't have systematic knowledge in computer science. I’ve been working on it but still fall short compared to others who come from a CS background. And this was also why I chose to work for some time after graduation. I treat it as an opportunity to work on my weakness: practice my programming skills, gain a better understanding of the production environment, broaden skill sets. I would have applied for PhD if I thought my profile was good enough. The company that I work for does not have much R&D going on, which kinda sped up the process of me going back to academia. The reason I’m aiming at the top CS programs is that I know what it feels like to do mediocre research at an average school(sigh on my my first Masters…  not saying that you cannot get to conduct exciting research at average school, but you’d have a better shot studying at a top school). I’m totally okay with working as a software developer if I don’t get accepted into top CS programs. But if I could feed my curiosity during PhD and work as a research scientist after graduation, that would be awesome. 

 

I have publications in top conference but that was not in the field of robotics&AI.  So I’ll need address that in this Masters program coming up. In the following 2 years, I’m planning to try to get a good GPA + strong references (preferably from profs with connections to top schools in US) + publish 2 papers in top conferences (hopefully more. I’m sure there’s exciting things I could do at the lab that I’m going to attend) + get decent GRE scores. 

 

Anything else I need to pay attention to? Or do you think this plan is not quite practical? 

 

I've only discussed with my close friends about it. It feels quite different to post it online Any thought would be appreciated.  Thanks!

 

[zliu224]

It's doable. But you need to be really determined and focused. It also means that the next two years could be the toughest year you would have before you get into a Phd program (let's not talk about after getting into a Phd program when you would scratch your head working on your thesis for 4 ~ 6 years )

 

I think for most top Phd program, they mostly care about your research experience. It doesn't matter the quantity, but the quality. I know many people with multiple publications get rejected from these schools (I assume you are talking about CMU, MIT, Berkeley, Stanford). Also, many people with 1 or 2 1st author publications at top conference get accepted. The time you would spend on these research as a first author will be significant. If you are lucky, you research could get accepted into those top conference at first try. If not, you will face the question of how to improve upon existing research. 

 

A good friend of mine suggest work on multiple  (but not too many) research projects at once, but have one or two that you are mainly responsible for. Whenever you have a chance, try to submit to conference. 

 

But there is also a question of whether you can find an advisor that would give you the freedom to lead research as a master student, since most of their responsibilities are for Phd students.

 

In the end, I think it really depends on the individuals. I believe if you are really determined, you can achieve whatever you want.

[heypeach]

@zliu224:

 

Thank you for your encouragement and concrete advice. 

 

I agree that the quality of research is what matters. The potential advisors that I've talked to, seem pretty nice and do give freedom to master students to lead research. (I guess it's a benefit of doing masters in Canada).

 

With regard to working on multiple projects, does it have to be closely correlated projects? Like planning & computer vision in robotics applications, which can be different aspects of one research projects. Is it still preferable if the two projects are less relevant? say working on computer vision in a robotics project while working on a gaming AI project. I've been thinking about finding co-advisors (one in robotics, one in AI). So I'll likely be faced with this situation soon. 

Edited March 19, 2015 by heypeach

[zliu224]

I would suggest concentrate one area (Others can disagree with me). Grad school like an applicant with a consistent profile and background. 

 

It would be better if two projects are related in one way or the other. 

[heypeach]

I would suggest concentrate one area (Others can disagree with me). Grad school like an applicant with a consistent profile and background. 

 

It would be better if two projects are related in one way or the other. 

 

Thanks. I appreciate your insights.