Please provide your feedback about the website at feedback@apc.iisc.ernet.in
Questions from participants of the ‘Why PhD?’ event held on 27 October 2022, answered by faculty members of the Division of EECS
1. Which route is better – BTech–>PhD or BTech–>MTech–>PhD?
Mayank: Procedure-wise, those who join PhD after BTech are required to do more course work – eight courses, compared to four. There is one semester of extra course work needed for those who join after BTech. Other than that, there is no difference in terms of the requirements. In some sense, you are saving one-and-a-half years. But I would say you are actually saving one year. MTechs come with some level of training on various tools and skills needed. BTechs don’t come with those trainings, but those trainings are also offered. IISc also offers an integrated PhD position, wherein if you have a certain number of credits, you get both Master’s and PhD degrees. This means that you don’t lose anything; you only gain an extra year, and that extra year can be invested in many things in future.
2. Is MTech mandatory for joining a PhD programme?
Mayank: No.
3. What are the expectations from an MTech student wishing to pursue PhD?
Please watch the initial 19 minutes of the ‘Why PhD?’ event here.
4. What is the difference between MS by research and PhD?
Chaya: MS by research is a research programme of a shorter duration. If you are sure that you want to get a taste of doing research, but you don’t want to commit for a longer duration, then perhaps MS by research is a good idea. In terms of the work that you do in MS by research, you typically either solve one problem or a couple of problems that are closely related and write an MS thesis whereas, in a PhD programme, it is more like a journey where you go from solving problems that are given to asking good questions yourself and becoming an independent researcher. So, it really depends on what you want to do, and what your goals are after your degree. If you feel like you want to try doing research, but you are not sure if a PhD is good for you, then you can enrol for an MS by research programme, which is sort of a good intermediate point, and if you have fun doing it, you can upgrade to PhD or finish your Master’s and then enrol for PhD.
Mayank: You can also request for conversion from MS by research or MTech to PhD. At least in IISc and some of the top institutes in the country, this conversion process exists.
5. What is the difference with respect to course work, duration, placement, and project in PhD, MS, and MTech?
Vishnu: If you look at the different departments, there may be some variations, but the course framework is very clear with respect to the MTech programme. Typically, you are expected to do roughly about 40 credits of course work. Each course may be 3 or 4 credits; they may be graduate or post-graduate level courses. Then, there will be a project component, which is typically around 24 credits. MTech students complete the course work in one-and-a-half years.
Mayank: Let me put this in this way: MTech course work – 12 courses; MS – 6 courses; PhD – 4 courses if you join after MTech, 8 courses if you join after BTech. Typically, you do 5 courses in one semester. Placement is the last thing that IISc students worry about. I hope this statement makes it clear what the placement opportunities are. Students in IISc think about many other things, and placement is the last think to worry about. Regarding the project, it all depends on areas, groups, faculty. It is completely diverse. It cannot be classified as PhD, MS, MTech. But typically, MTech projects are fixed duration projects, MS projects are also fixed duration projects in some sense, whereas PhD is a training programme, it is not a project per se. It all depends on how fast you can demonstrate your ability to do independent research.
6. Can we know about part-time PhD?
Chiranjib: The only way we can enable part-time PhD is through the external registration programme. Here, there are certain rules, which are publicly available. The summary of the rules are: the company should give you leave for one year so that you are physically present here. There should be an advisor from the company side. After one year, you can go back to your company, and contribute both to your company and to your thesis.
7. Do I have to pay huge fees if I get admission to the external registration programme as a PhD candidate from a Government organisation (when the parent organisation does not fund the fees)?
Mayank: The organisation has to sign for funding your fees. This fees is higher than that for the regular PhD candidates, who pay from their stipends.
8. Can you explain the advantages and disadvantages of doing part-time and full-time PhD?
Chiranjib: The only way to pursue part-time PhD is through the external registration programme. As I have seen over the years, most of the ERP students are senior, that is, they have a family and also a job. So, the good thing is that they are much more focussed than our usual students who are full-time PhD students. However, for our full-time students, it is perfectly alright to say, “I have no idea what I am going to do research on. Let me read lots of papers, lots of books; let me try to find problems and try to solve them.” We welcome that; that is actually the best training you can have, because you should not only have a good problem to solve and lots of papers to write during your PhD, but you should also come out as a well-rounded person who is exposed to many things in that subject. For that, our full-time PhD has a big advantage. But, for ERP, your company would also expect you to do some work because they are paying you a salary after your first year. Often, most companies do not leave the candidate alone since there are some expectations. So, unless a candidate is very focussed, unless you have a good mature problem and a good direction, that can become a challenge. This would be a very high-level advantage/disadvantage. So, a one-line summary would be that a full-time PhD allows you to explore many topics, and I personally think this is absolutely essential for somebody with a PhD degree, whereas in a part-time PhD (the ERP programme), because of the lack of time, it will be good if the candidate is focussed. However, there are several examples in our department of ERP candidates who have excelled, and they have written excellent papers.
9. How can we apply for PhD at IISc?
Chiranjib: Currently, we are preparing ourselves to introduce some candidates in November. So, you can still apply. Go to IISc website, you will see admissions and there are clear instructions to fill up certain forms. When you apply, each department—according to the criteria the department chooses—shortlists a select set of candidates. The candidates will then be called for an interview or a written test, and this varies from department to department. For example, in CSA department, if you are shortlisted for an interview, you will be asked to choose a stream – either computer systems and software or intelligent systems or theory, and appropriately, there will be questions asked, and we often stress on foundations. It is perfectly alright to say, “I do not know advanced topics; I only know foundations, but I am pretty strong in that.” Those are the type of candidates we are looking for. So, we check for basic curiosity and mastery at the undergraduate level. If you already have a good Master’s degree, it’s even more of a plus. So, this is the way we select our candidates.
10. How do you select candidates for PhD at IISc?
Please watch the initial 19 minutes of the ‘Why PhD?’ event here.
11. Can I pursue PhD after two years break from my studies?
Mayank: Yes.
12. If I am planning on a PhD after 2–3 years of industry experience after MTech, what points should I keep in mind? How can I update myself in the desired field of research while doing an industry job?
Chiranjib: We all have our own views. My research area is in machine learning. So, my answer will be tuned towards that, from that perspective. What I have found is that often, in most of the jobs which are offered here, skills such as proving theorems and reviewing papers are not utilised; hence, in two years’ time, some of these good, strong, academic foundations tend to get lost. That is one case of worry. So, jobs that especially have a large software development component actually cause decay in your academic credentials. That can be very bad news for PhD. However, if you are in a job where you are constantly reading papers, supplementing them – then, that is exactly the job profile which would definitely motivate you to do PhD. It prepares you well. So, my suggestion would be that if you are planning for a PhD after 2 to 3 years, please keep in mind that you should keep yourself abreast of your GATE syllabus and mathematical subjects. You should have familiarity with that. If you are already doing a research-related job, you are very well fit to walk into a PhD programme. Having said that, there is definitely merit in having an industry exposure because that is where there is real excitement.
13. What are the opportunities to do a PhD while working simultaneously in industry?
Mayank: First of all, PhD is a full-time engagement. There is no PhD possible while working in industry except through the external registration programme (ERP) or the sponsored programme that we have. And, even in that case, your engagement on a PhD problem is priority number one. This is also a myth that PhD is a project, and so, you can take up one more project and you can continue working on your industry job. This is not the case. If you are working in industry, either you leave that job, come for a full-time PhD, stay in IISc for a longer period, or if you want to be in your industry job and if your employer supports it, you can enrol for the ERP programme; but even there, you have to commit for the entire course work and after the course work, your administration should give you enough bandwidth and time to work on your PhD problem. Usually, when we sign papers, we expect that the industries commit their employees to be available at least 50–60 percent of their time to work on the PhD problem. Your PhD problem will become your full-time job, and your industry job will become your part-time job.
14. What are the opportunities for a post-doctoral programme at IISc Bengaluru after a PhD in the field of biomedical signal processing, artificial intelligence, and machine learning?
Vishnu: There are different types of post-doctoral programmes that are being offered by IISc. There is a National post-doctoral fellowship; there are also Institute-level fellowships such as the C V Raman post-doctoral fellowship and the Institute of Eminence (IoE) post-doctoral fellowship. Those are all schemes, but you have to first identify a faculty who might be interested. I would recommend you to get in touch with faculty who are working in the field of artificial intelligence (AI) and machine learning (ML) who might be interested in your background. In the department of electrical engineering (EE), we have many faculty who are working in the field of biomedical signal processing, and one of them is Professor A G Ramakrishnan. Professor Sriram Ganapathy is working on speech processing. Professor Prasanta Kumar Ghosh and Professor Muthuvel Arigovindan also work in related areas; you can consider reaching out to them. There are multiple faculty who work in this area, and I would recommend you to get in touch with them.
Sudhan: In the department of electrical communication engineering (ECE), Professor A P Prathosh is working on AI and ML for biomedical signal processing.
Chaya: In the department of computer science and automation (CSA), intelligent systems is an entire research stream, and we have several faculty colleagues working on AI, ML, and related problems. We also have some post-doctoral fellowships like the IoE fellowship. You can check out the department webpage – the information is there. If you think it is suitable, you can apply.
Mayank: In the department of electronic systems engineering (ESE), you should contact Professor Hardik J Pandya and Pofessor Chetan Singh Thakur who work on biomedical signal processing and biomedical topics using AI and ML. You can also contact Professor Shayan Srinivasa Garani who works on signal processing and applications of AI and ML in signal processing. More than 30% of EECS is now in AI/ML and its applications in various fields.
15. Regarding students who want to take up roles in nanoelectronics offered by multi-national companies such as Intel/TSMC/Samsung – what will hold them back in India after a PhD? I am talking about working with leading-edge technologies such as transistors.
Mayank: Whether you stay back in India or you take up positions abroad, that’s completely your choice. IISc prepares you to an extent that is parallel to any best institute in the world, and when you trained to that level, you will have opportunities from around the world. Then, it is completely your call to decide whether you want to stay back in the country or you want to take up a position elsewhere, depending upon the opportunities. Many of us spent a good time in the United States or other parts of the world, and decided to come back for our own calling. So, at the end of the day, it’s your own calling – what you want to do, and where you want to do what you want to do.
16. If a person decides to stay back and work in academia in India – where can he/she get funds to conduct independent research (other than institutes that are a handful in number)? What would even drive him/her if not for industry–academia collaboration and facilities?
Mayank: The drive is completely yours. There are several funding opportunities. Among the researchers and often faculty in tier 2 institutes, there is a myth that the funds are often biased and that they are available only for top institutes. Many of us have been in these committees, and we can tell you that these funding allotments are highly competitive, and they go through a rigorous screening and evaluation process. In that screening and evaluation process, nobody looks into your affiliation. Of course, there is often a correlation between your affiliation and how good your curriculum vitae (CV) is. People often miss out the CV part and often, unfortunately, correlate with the institution. But the fact is that one would look into the proposal, and one would look into the CV. The CV tells of the promise of the individual to deliver the promise or the confidence that the funding agency gains in terms of the individual being able to deliver the promise given in the proposal while the quality of the proposal tells how much advancement in the state-of-the-art will happen if that proposal is pursued or executed. So, if you have written a good proposal, and if you have carefully built your profile, by taking up every opportunity that comes in front of you from the initial stage, after a certain time, you will not feel that your proposal has been kicked out. You will have funding available if you are doing well and if you are writing good proposals. It is an exponential path. So, if you do well, you will do further well and you will have more and more opportunities. Now, if you get into this exponential path, you will have the drive.
17. What happens to research works/papers after they are published? Who actually uses them?
Vishnu: When you do research, whatever be the research, you always stand on the shoulders of giants. You are always looking at what others have done, and that gives direction to your work. In that sense, it is going to be useful to future researchers. You might have used some type of experimental technique or you would have proved a theorem in a particular way, and somebody might use that result as an inspiration to find out something new. All published research will have their own uses.
Sudhan: Once you publish a paper, assuming the paper is not patented, or even if you patent it and then publish, this work can be used by the community. In a specific area, there will be more than a thousand to million people who will be working in the same direction. If your work is very important and significant, then they will cite your paper once they use your method. Then, you will get credit based on how many researchers cited your paper, and that will add value to your CV and your publications; that will also help you for your promotion and when you are looking for a job. You can benefit in many ways if your published paper is very good.
Chaya: Whenever we publish a paper, we are basically advancing the state-of-the-art and knowledge. We either prove a new theorem that gives us new knowledge, or if it’s not a very theoretical paper, we show a better way of doing things that is more efficient than how we currently do things. If you are advancing the state-of-the-art in a theoretical way, then you are basically giving other people more tools to build on top of that. You are contributing to expanding knowledge in some way. If you write a more applied paper, there are also other ways of it being used. The result can actually get deployed in the real world and be made use of in the industry. For example, in computer science, if you come up with new algorithms for very relevant problems or in cryptography, if you come up with cryptographic schemes – today, there is a lot of scope for them to be deployed. This was always true. There used to be a time lag between when you make a certain contribution and when you can actually see that implemented in society. Today, I think that gap is reduced, and you can actually live to see your science making changes.
Mayank: Often, students don’t know what to do with a paper, and often, students are misled in terms of publication as a requirement. What is a paper? A paper is basically your communication about the advancement you have done in a certain field of science or engineering. That advancement is based on a problem statement that you pursued, and you tried to solve. Basically, your paper talks about a certain way of solving a problem. There may be many ways of solving the same problem, there may be many views on the same problem. You are communicating your way of solving that problem or elaborating that problem or even understanding that problem better. Now, how does the problem originate? The problem may originate from something which is pertinent within industry, something which is highly in demand within industry. Then, through peers and the community and through other papers, you get to know a particular problem that the world is trying to solve. That is one. The second is that a problem could be fundamental in nature, say, understanding certain aspects of nature or the universe. It could be physical sciences, it could be biological sciences, it could be mathematical sciences, chemical sciences… So, basically, you are trying to understand a certain property or uniqueness of nature, which has not been explained or elaborated before. That’s also a problem statement. The third type of problem statement could be that a community or a group of people have come up with certain predictions for tomorrow – that tomorrow the world will be of this kind in terms of science and technology, and you are trying to basically develop the idea further by either pointing out certain problems that the world needs to address or by working on problems that other colleagues would have pointed out that this field has to evolve or develop. You have to work on these problems, for example, quantum technologies. These are areas where the problem has originated from the vision that the community has projected. Now, this gives you the problem. If you have solved that problem, the solution or the paper goes back to the same community which has basically given that problem. That community could be your academic circle, it could be your industrial circle, or it could be your other scientific community or colleagues who were posing those problems, or who were parallelly working on those problems. Either it can result in advancing the state-of-the-art of the technology or in an understanding of nature or science. When people cite these papers, you also get noticed as a result of that. Fundamentally, the paper goes back to the community which has posed the problem to the community.
18. Any suggestions on how to work on our dream projects during the current BTech itself?
Chiranjib: In IISc, there are opportunities for doing internships after your BTech. Unfortunately, you don’t have a good, structured programme. Often, you can look up Professors’ websites and see if the things which you want to learn about and want to do your dream projects about is aligned with the Professors’ interests. Do write to the Professor. If you don’t get any response, it is not that the Professor is not interested but it may be that the Professor does not have any positions open. If there are positions open, they will definitely get back to you. We welcome young minds.
