Search This Blog


Wednesday, April 8, 2015

How to Choose An Engineering College (2015)

This is my second blog article on admissions related issues. I wrote one with exactly the same title long time ago in 2009. I then expanded upon it and wrote an improved version, but did not share it as a blog article. Instead it remained on my website. So it didn't reach as many people, since I notice that a lot of people still access my 2009 article. So here is the expanded version, largely same as what you would find on my website.

The reason I wrote this in 2009 and again in 2011 is rather simple. I get lots of emails on my views on different engineering colleges, and request to compare this with that. I am not in the habit of comparing things based on what I have read on social media or what I have heard on the corridors. And hence in most cases, I have to decline any comments. And I thought instead of just saying that I have no opinion, I could give them an algorithm to make some progress in coming to a decision. And hence this article.

First of all, which is a good college. Yes, I know. The highest package. Sorry, if you think so, and you are convinced about it, this article is a waste of time for you. Please stop reading. Go to any of the infinite sites where you can ask questions about packages and you will get answers that you can completely trust.

Let me suggest an alternative. It prepares you to achieve your goals. And, of course, your goal could be to be rich. But you don't become rich by having large paycheck in the first month. You become rich by having a long paying career. The first job is no guarantee of a successful career. And placement statistics anyway are most unreliable. So may be you should be looking at how alumni are doing 5-10 year hence rather than how final year students are doing. But then that statistics is even harder to get.

So let us look at factors that are most likely going to lead to a successful career. These factors are:
  • Your passion and interest in the area you are working in (that is why it is important that you don't choose CS or ECE just because everyone else is doing so, but think of what excites you)
  • Your preparedness as far as knowledge and skills are concerned (this is why the quality of education is so important)
  • Your ability to keep learning lifelong (this is why good faculty is absolutely important - to not just teach you some technical stuff, but tell you how to learn yourself)
  • Your soft skills, attitude, ethics, etc. (this is best learnt at school, but a good college would improve upon this, particularly a residential campus where you learn many things from hostel activities)
  • Your network of friends (this is where a good college which attracts more good students will help), and
  • A huge amount of luck.
This article is not about choosing a branch, though I would like to add that if you do not have a particular interest in a discipline (and most 12th class students don't - it is ok), then prefer a good college over a popular discipline.

And this article is not about luck. I can just wish you all a huge amount of luck.

So let us focus on the other four factors.

The most important factor is the quality of faculty. It not only helps you in learning how to learn lifelong, but also ensures that you pick up adequate knowledge and skills from the program. How do we know which college has a better faculty than the other. Well, visit their website, and look for the following information:
  • Number of full-time faculty members. Please make sure that you read the details, and find out who is a full-time faculty member, and who is a part-time or adjunct faculty member. The important parameter is faculty to student ratio.
  • Their qualifications. How many are PhDs. Where did they do their PhDs. Similarly, how many faculty members are having MTech qualification. Where did they do their MTech.
  • If a significant portion of faculty received their highest degree (whether PhD, MTech, or BTech) from the same college, then that should raise some alarms. On the other hand, faculty members having a lower degree from the same college, implies that they value the place enough that they returned back to the same place after getting a higher degree from elsewhere.
  • If the highest qualification for any faculty member is BTech or MCA, then be alarmed. Top places will only higher PhDs. Good places may have some MTechs. But if colleges are hiring BTechs and MCAs for teaching courses, it means that they are not able to attract enough good faculty, and that should be a cause for concern.
  • What are faculty doing. Are they teaching three courses a semester or two. Are they doing at least some bit of research publications.
Of course, people will argue how important it is to be a PhD to be a good teacher. And I have no doubt that there are some excellent teachers who are not PhDs, and there are some lousy teachers who are PhDs. However,  there is no doubt that places which have a lot of non-PhD teachers have them because they failed to attract PhD faculty, not because they just hired great teachers, and it so happened that many of them were MTechs. And if you look at the background of those MTechs, it is quite unlikely that you will find many of those MTechs from IITs, IISc, and other top institutions. And, in general, PhD from a good university would have a higher chance of being a good teacher than an MTech from a tier two college.

Another factor that gives an indication of the quality of faculty is the research output of the institution. I believe that there are good researchers who are not good teachers, and similarly, there are good teachers who are not good researcher. However, in general, faculty members who are actively pursuing some research interests would be current on the subject and would have a deeper understanding of the topics. But more importantly, research flourishes when there is an institutional support for it. If faculty members are doing research, then it shows that the college management is serious about the quality of education. Research can be measured by the following parameters. (You may need help from someone in the academic community to really understand these parameters. Go ahead and ask whoever you know.)
  • Publication by faculty members. Higher preference should be given to peer-reviewed journals and conferences of high quality and reputation. Then one should also look at local conferences. At least someone is putting some effort in the right direction.Unfortunately, there are enough journals today where one can publish by paying. There are also online sites where anyone can upload a paper. So it has become extremely difficult for anyone but a researcher in that area to figure out the quality of research by looking at the webpage.
  • Sponsored research projects by various funding agencies like Department of Science and Technology (DST), AICTE, Dept of Information Technology, etc. Normally, research projects are peer reviewed and thus presence of such projects indicate that peers think highly of them.
  • Any industry interaction in terms of research projects or consultancy. Working with industry requires confidence in one's abilities. Industry isn't in the business of charity. Also, industry interaction often would lead to good internships and employment opportunities for students.
  • Do they invite several researchers to give seminars. Do they organize workshops and conferences. This shows that the place is active and has energy. 
 Now, let us look at the second factor for a successful career. That is, level of preparedness or the knowledge and skills learned. This will, of course, depend on quality of faculty, which we have already discussed. But it requires a couple of other parameters as well. Most important of them is the curriculum. Some of the things to look for in the curriculum are:
  • How many courses do they teach. Unlike the conventional wisdom in India, I believe that the college that teaches you less is a better college. It means that they do less spoon feeding, and give you more space to grow and learn. There are surely exceptions to this general trend, but by and large colleges will try to teach you more, if they know that they are doing a poor job of teaching, and hope that if they try teaching you lots, then perhaps in some courses they will be able to teach you something.A good college may have 40-45 courses in the curriculum, while a poor quality college may have more than 50 courses.
  • How many electives are there in the curriculum, giving flexibility to the students to learn what they are interested in. Many colleges may have slots for electives, but they treat that slot as their choice to offer a course. So they won't offer three courses, and ask students to chose one. But instead they will offer one course of their own choice (basically for whatever course they can find a faculty). Also, in most colleges, the curriculum will only contain professional electives, but no open electives.
  • Do they have enough number of humanities and social science courses (at least 10 percent courses). One cannot be a complete engineer without understanding economics, sociology, psychology, etc.
A lot of learning happens outside the classroom, and hence a residential institute should be preferred over a place where all are day scholars. If there is a mixed system (that is, some live in hostel, and some are day scholars), it is still better than fully day scholar since even if there are some students on campus 24x7, it would have facilities that even day scholars can use when necessary. You won't have all labs close at 5 or 6pm. The library is likely to be open late. Indeed, one of the parameters to look at while understanding the quality of an institute is whether they allow access to their facilities for long hours, or are they only from 9 to 5 on weekdays.

Another important criteria is the autonomy of the institute. Can they decide their own curriculum. Typically, universities (including deemed-to-be-universities) can decide their own curriculum, and in general I would strongly recommend universities over affiliated colleges. Teaching someone else's curriculum is demotivating for teachers. If they do not have much stake in the curriculum, it would also invariably mean that exams are also conducted by someone else (by the universities, except for some "autonomous" colleges), and that means students don't care for the classes and teachers. This can not be conducive for lifelong learning, not even for immediate learning. But, of course, a vast majority of engineering education happens in such affiliated colleges, and most of it is poor quality. Of course, this is generalization, and certainly there are some affiliated colleges which are doing a decent job.

Fancy infrastructure is not something that impresses me, but yes, they should have all the necessary labs, good Internet bandwith, WiFi access so that you can use your own laptops and other devices anywhere, a good library with lots of reading spaces, lecture rooms without a projector is like living in dark ages, adequate sports facilities, etc. (Caution: Some of the engineering colleges would have all of this and more, but would not have faculty. Look at infrastructure only after you are convinced about the faculty and curriculum, etc.)

To ensure that your peer group is strong (since so much of learning will happen outside the class room, and your career will be helped by a good peer group), one may look at the closing ranks of the admission test (like JEE). The other possibilities are various rankings that you see in the media. Well, I would just ignore all the rankings. I have seen such stupidities in these rankings that I would not want them to influence your decision at all. But then there is a problem. How do you compare two institutes who take admission through two different criteria. Well, my advice, don't use this criteria in such cases. Of course, this criteria will also discriminate against some new institutes who may have suddenly become very good (like IIIT Delhi in comparison with other Delhi institutes), so take this with a pinch of salt.

Other things that you might want to look at:

  • Quality of the website. Do they give you all the information that you need to evaluate them. If they don't give some data, assume the worst. In this age, if they are not serious about even this level of publicity, how serious will they be about your education. How will they attract industry for placement and internships.
  • What fraction of graduates go for higher studies? This shows that they are not sick of their education. And also, the higher education is becoming increasingly important for success. So the chances are that a greater fraction of their graduates will be successful in future.
And things that you should not look at:
  •  Placement data (it is completely unreliable and has no prediction value)
  • MoUs with foreign universities
Then there are things which do not really matter in terms of quality, but could be important for you. Feel free to factor them in, and indeed they are important. One is Geographical location. Many people have preference to stay close to home or away from home, in a similar cultural environment or in a similar weather condition, etc. This is fine. The other is finances. If the two places you are considering have very different costs, then one has to look at whether those differences are worth the extra cost. And it is never going to be easy to take a call on that.

I will perhaps keep editing it as I think of more issues.

Tuesday, April 7, 2015

My 2015 List of Recommended CS/IT Programs Outside the IIT System

JEE Mains is over for a majority of students who gave the offline version, and long before the result (which will take three months for including the weight of board marks), a lot of people are already discussing their options. I had written two blogs in 2011, one for deciding which program to opt for in which IIT and the other one mentioned a small list of institutes outside the IIT system which are good and which have something unique about them. The two blogs are in the top few in terms of readership of all the blogs that I have written in the last five years of my blogging life. But some of the information in those two blogs has become outdated and I thought I will write a bit about the non-IIT places which have strong IT/CSE programs.

It is much easier for me to write about these programs now than it was in 2011. The reason is that I haven't traveled to most of the departments that are perceived as good for many years, and as I wrote in my blog in 2011, I wouldn't want to include any institute unless I have good information about it. So, my list is going to be rather small this year. It has three institutes from the older list:
  1. International Institute of Information Technology, Hyderabad
  2. Indraprastha Institute of Information Technology, Delhi
  3. Birla Institute of Technology and Science, Pilani
I had mentioned in 2011 that IIIT Hyderabad is the best institute for IT education outside the older IITs, and that IIIT Delhi was a brand new institute that had come up really fast within a short 3-year period then. Now, four years later, my impression is that IIIT Delhi has grown in quality so much that it can be compared with IIIT Hyderabad. IIIT-H continues to do well, continues to attract good faculty and top students from JEE. They continue to have a great crop of research students. IIIT-D, on the other hand, suffers from lack of brand recall amongst the students and parents. However, a look at their website will tell you that they have built a truly great institute. The faculty quality is awesome. The research output is great. The curriculum has all the right ingredients with a good bit of flexibility. The brand new campus infrastructure is excellent. And given all this (and good placements too, though I personally do not really want to compare those numbers), I think it is only fair to say that they have caught up with IIIT Hyderabad.

Indeed, I am so impressed with the growth of IIIT Delhi, I have decided to join the faculty of IIIT Delhi from the next academic session. (Here you go, take everything I say on this blog with a pinch of salt. I am totally sold on IIIT Delhi. That could also mean that I may have some biases.) I will write a separate blog article on what all is great about IIIT Delhi some time in near future.

BITS Pilani is a place that I admire a lot. Every time I go there, I learn something new about how to be student friendly. I admire their flexibility, particularly their dual-degree programs, their admission process (BITSAT is any day better than JEE), and so on. One of their strengths is the diversity of people on the campus. It is not an IT focused institute, and has a large number of departments. The interaction with people from different disciplines can only help the students. Of course, they have a fantastic brand name and a very active and helpful alumni group, which is a huge plus for anyone joining BITS. But purely in terms of their IT/CSE strengths, I would rate them after the two IIITs. (You, the student, should be able to combine the pluses and minuses and decide the relative ranking for yourself. Everyone's mileage will vary.)

Now the newcomers. I have only one to add. And surprisingly, it is not an engineering college. It is a new university, which prides itself in being one of the finest universities for liberal arts education. Yes, I am talking about Ashoka University. It started just a few months ago, and the first under-graduate batch is in its second semester. They have several majors from humanities and social sciences, but they also offer a major in Computer Science. Of course, it is a three year program as of now, and not a BTech degree. So it is perhaps not a good fit on this page. Also, they have not been able to hire many faculty members. So there are issues. However, considering the quality of people whom they have recruited in different disciplines, there is absolutely no doubt in my mind that we are seeing emergence of a truly world class university. The quality of students they have attracted in the very first batch is fantastic. The infrastructure they are building is great. The best part of the program is that you get a very liberal education. All under-graduate programs are supposed to be broad based, and in any engineering college when you do a degree in Computer Science, you will be exposed to other engineering disciplines, other science subjects, some humanities and so on. Ashoka too will give you a broad based education, but those non-CS courses would be mostly from liberal arts and may be a few from science, and I think this kind of broad education will be really useful and interesting for a whole lot of students.

Let me comment about some other institutes as well. While I have not visited DAIICT for many years (surprising since I have been to Ahmedabad dozens of time in this period, and I have many friends in DAIICT), I am told that the faculty student ratio has declined a bit in the last 4 years, but not to any worrisome degree. A few older stalwarts have retired, but the new faculty is continued to be recruited carefully. A real big problem is that there is tuition control by the state government, which means that the resources at its disposal are less than what they used to have, or what other similar institutions have. The admission is also regulated by the state government. They also had a change of leadership recently. We will need to see how it pans out for them. But overall, I am confident (even without a visit) that it remains a good place for IT education. If I visit them before the admission season, I will add more comments here.

IIIT Allahabad used to have a serious leadership issue. They now have a new leader (since last one year). I hope things are improving there.

Prof. I K Bhat moved from NIT Hamirpur to NIT Jaipur. And there have been several improvements in the department. My ranking for NIT Jaipur has improved as a result, though I wouldn't put them in the same category as the first three places.

I may add more information in the coming weeks.

Monday, April 6, 2015

Net Neutrality

Came to know through this link that there is a proposal (or actual offering) to provide access to some sites for free. Airtel will charge those sites for data usage instead of the users. So Internet users with this plan get to surf these sites for free. Of course, users have an option to pay and then surf all the sites in the world. Isn't this a great plan. Most Internet users would visit only a handful of sites, and if all those sites are included in the plan, then it essentially provides free access to most customers. That appears to be a great thing to happen.

But let us fast forward this a bit. If a service provider like Airtel is able to have a deal with 100 top apps that they will pay for the bandwidth instead of the user, and if any popular app resists such a deal, the service provider can throw in that package anyway on its own cost since it is making enough profits from other 100. a vast majority of the users will seek this package and now this vast majority has no way to access any other site.

These customers may not mind that situation, but is this a healthy situation for the country, for the society, for further innovation and so on. It is obvious that this will hurt innovation. Any new company that wants to have its app distributed and want people to access its service will have to have the same deal with all service providers in the world who have such packages. The service providers may refuse to have such deals, which means that these innovative apps are not accessible to a vast majority of customers. And even if the service provider agrees to have a deal, the cost of innovation has just gone up substantially. This simply can not be a healthy situation for the society at large.

Will this even help those customers who are happy with that limited free access. To some yes, but not to everyone. If it is an e-commerce site, would it give away money to the ISP as charity. Of course, now. It is going to take that cost into account when it decides its prices. So the customers will end up paying anyway.

Isn't this similar to reverse charge calls or toll free calls where the receiver has agreed to pay for the call. If we can have that system in telephony, why not in Internet. Unfortunately, the system is not similar. If a telephone company offers you a service which allows you only free calls to a few thousand numbers that the company decided, would it be useful. In the Internet world also, if we had a system which allowed the site to subsidize my access to them without compromising or reducing my access to anyone else, it would be perhaps alright. But then this decision would be of that site, and not of the Internet service provider. (I feel uncomfortable with this, but have not been able to come up with a good articulation of that discomfort.)

And that is really the whole idea behind net neutrality. The network provider should be an honest broker between the users and the sites. It should charge for its services in a uniform way, treating all bits as equal. It can vary charges on the basis of technology, number of bits, speed of bits, time of day, day of week and so on, but not on the basis of source or destination of the bits, since that would lead to stifling of innovation in the Internet.

Of course, this is not the first time an Internet service provider has tried to differentiate between sites. We already have the controversial promoted by facebook. We had another attempt earlier of charging extra for VoIP services. And with TRAI releasing the discussion paper on Over-The-Top (OTT) services, the debate has started in India regarding the extent to which net neutrality should be maintained here. And I do hope that TRAI decides in favour of a strict version of net neutrality.