I joined the Phd program in the Electronics and Communication department in July 2017. I am interested in Image processing and machine learning including the latest techniques like deep learning. I spoke to three professors and have recorded my observations below:
Works in using deep learning for surveillance and driver assistance related applications
Famous guy with stellar publications. Spoke to him and he told me that he works in a particular manner and he expects students to adhere to it. I spoke to one of his students and he told that he gives the research problem and also the direction to follow. This was the field that interested me the most. But he does not give freedom to students to select their method or way of working.
Works in Medical imaging
Defines the problem statement but allows the students to select their method. Strict with very high expectations. Somewhat crazy, can behave erratically at times.
Works in deep learning techniques for driver assistance applications and computational photography
Brilliant researcher but an absentee prof.
I was confused between Prof's A and B. The work done by Prof A interested me a lot but the fact that he doesn't give freedom put me off. I selected Prof B as my Phd advisor. Incidentally, I have not yet started my work. I spoke to Prof B. about using deep learning for medical imaging but she always puts me off with vague answers. She selected a topic for me which doesn't involve deep learning in the foreseeable future without asking me. I told her once indirectly that I would like to use machine learning techniques for my work and she gave a very vague answer. So I asked her directly and her words were that "You can have it as an appetizer but it will involve using conventional techniques." So, I decided to dig into the work that has already been done and found that the deep learning aspect has already been implemented by a Masters student who has passed out. She has given an almost similar project to another student who indicated to her that people are taking machine learning based approaches to solve the problem. The prof told her not to look at those approaches.This led me to the conclusion that she is not really interested in letting me do work with any machine learning techniques. Furthermore, she tried to tempt me saying that she is expecting funds for another project and she will make me a manager for it and pay me for it. My concern is that should this happen, it will take away time from research and will greatly slow down my Phd. I am not liking the way she is not willing to give me the kind of work that I want and her wanting to exploit my industry experience for other things which will not help me. But I am not sure if he will reconsider me since he is basically a tough guy. I am extremely confused and don't know what to do. Please advise me.
You are only 5 months old and should be alright changing to a different sup since you have not invested too much time in your project.
So, you have decided that you want to ditch Prof B for Prof A? My question is how close are Prof A and B? Do they collaborate? If yes, Prof A may be highly unlikely to take you on as a student so as not to offend prof B. If no, then you may wish to ask Prof A's students if he has funds to take in students. If yes, then you can speak to Prof A detailing the areas of your interest and what you can contribute to his group. If you have scholarship that is transferable to this new group, you stand a good chance of getting an ok from him.
But are you sure that not being able to choose your methods is an acceptable tradeoff for you to learn machine learning techniques? Just want you to think deeply about it so you don't revisit this issue 2 years down the line.
Thanks for your reply. Yes, I have decided to ditch Prof B and join Prof A if he accepts me. I don't have any scholarship except the one the college gives me, so it should not be an issue. I have not started work yet, so that again should not be a problem. As you have rightly pointed out, I selected Prof B since her students told me that she defines the problem statement but gives you the freedom to choose the methods. But as soon as I started interacting with her and indicated my interest towards machine learning, she started pushing me away from it. She only wants me to use a tool with old fashioned methods. So it looks like I will get stuck with medical imaging which was my second choice and not have the freedom to select my methods.
Better early than late. I'm not encouraging you to act too fast, but if it does,'t feel right, you should act fast. At this point you are not too engaged with your research topic, and you can make the switch 'gracefully', without making you or you current supervisor look too bad.
Can I put on a different slant to this? My research was going in one direction (network analysis) and my supervisor pushed me into another (political history) - I fought this but ultimately accepted a lot of what was said to me.
I am so grateful about it now.
Because my sups was against my original method, the struggle made me really strong on why I wanted to use it and really back my arguments.
On the other hand, me capitulating in part on the suggested methodology actually added strength to my own intellectual trajectory/thesis/material etc.
So I'd say some conflict with sup may be actually good. Hate to say this because being in a battlefield is not fun, but choosing a 'yes' person is not always the best choice for one's intellectual journey.
Masters DegreesSearch For Masters Degrees
An active and supportive community.
Support and advice from your peers.
Your postgraduate questions answered.
Use your experience to help others.
Enter your email address below to get started with your forum account
Enter your username below to login to your account
An email has been sent to your email account along with instructions on how to reset your password. If you do not recieve your email, or have any futher problems accessing your account, then please contact our customer support.
or continue as guest