We have seen a rise in the number of AI-related courses offered by premier Indian institutes in the last couple of years. Many IITs have even started offering undergraduate courses in AI.
These institutes are often a yardstick to measure the quality of education, the rate at which they innovate, and more. They have been the torchbearers for introducing advanced studies to the Indian education landscape. However, when it came to taking analytics seriously as coursework, we may have lagged behind.
Let us take a look at few programmes that were launched since 2019:
- IIT Roorkee partnered with US-based CloudxLab to launch an Advanced Certification Course on Deep Learning
- IIT Hyderabad becomes the first Indian institute to launch a full-fledged BTech program in AI
- IIT Hyderabad also started a PhD program in its Department of Artificial Intelligence
- Indian Institute of Technology Guwahati (IIT-G) launched four programmes, including a Master’s course in Data Science
- The Indian Institute of Technology, Ropar launched a new MTech course in the field of artificial intelligence and machine learning and will intake 15 students for the first batch
- Indian Institute of Management, Ahmedabad (IIM-A) is offering a new 16-month post-graduate diploma in Advanced Business Analytics (PGP-ABA), from the academic year 2020
Along with these programs, IITs from Kharagpur and Madras have set up their Center of Excellences, and have partnered with companies like Bosch to promote AI.
So far, so good, but is this enough? The answer is both yes and no.
There are 23 IITs and 20 IIMs in total, and these institutions are considered to produce one of the best talents in the country.
But IIT Tirupati has no data science program either in UG or PG. In the case of IIM Visakhapatnam, like many other IIMs, it has no PG course dedicated to data analytics.
Now the argument does come around to “data science is in its infancy”. There is no doubt about that. That argument probably bodes well with those that play catch up with the latest trends instead of leading them. In the Indian segment especially, the top IITs from Madras, Bombay or IISc Bangalore and the top three IIMs always come in the news and appear on world rankings, but what about the 20 other IITs and IIMs?
The brand IIT and IIM are of immense value, and when a new institute is gifted with such a brand, it is obvious to expect standards that have been set by its predecessors. But only recently, institutes like IIT Bhilai have started offering data science programs. It is never too late, but our initiative has not been that early either.
Where The Struggle Lies
Data Science is the hottest topic, both in academia and in the corporate world. Unfortunately, it is also burdened by hype. There are only a handful of people who can convey this multidisciplinary subject concretely. So finding these people and making them teach it in a way that helps the students work on real-world problems is quite challenging. While the hype around data science draws more people towards it, many are quickly disappointed when they figure out that they will have to learn high school maths like linear algebra and tricky statistical concepts.
Finding the right talent has been a major hurdle for both academia and industry. Those who are good at it are already communicating through blogs, forums and YouTube. So, these institutes have the onus to offer well-curated coursework that lives up to the hype. Because, today in the age of the internet, you cannot get away with mediocre content just because you are branded IIT or IIM.
What Makes A Good Course
Though data science or AI on a broader scale, is still relatively new, the foundations, however, are rooted in century-old mathematics and statistical theories. For any university to start an AI department, they do not have to look far. It is always the combination of the departments of computer science, IT and mathematics. The above picture illustrates what Stanford data science undergraduates would have in their curriculum.
A large number of programs related to “Big Data” have been started in colleges and universities around the world, more so in the US, in the last two to three years. Other than ‘Data Science’ they often go by names like: ‘Data Analytics’, ‘Big Data Analytics’, ‘Business Analytics’, ‘Data Mining and Knowledge Generation’, ‘Applied Statistics’, amongst others.
These programs are typically post-graduate programs of 2-4 semesters’ duration leading to certificates, diplomas or MS degrees.
These programs typically provide mid-level and high-level managers with a basic understanding of the vocabulary, techniques and tools of data analysis and sensitises them to the power of data analytics to improve business performance from a ‘bird’s eye-view’ level.
Does this mean that ALL these institutes are clueless? No. On the contrary, IIT Kanpur does seem to stand out. They have recently issued a proposal on how a data science coursework should look like and have even incorporated them into their programs.
The program that is being proposed differs from other programs across the nation significantly.
Along with an introduction to machine learning techniques, this course work is designed to offer the following domain expertise to students:
- programming, algorithms
- operating systems, databases
- signal processing
- probability and statistics
- inference, optimisation, statistical simulation and data analysis, sampling theory
- management decision analysis
- decision models & game theory
With top institutes like MIT offering their courses for free online and not to forget the rise of inexpensive MOOCs, it is a great injustice to students to undergo four years of undergraduate courses which offer no significant depth than what is available for free on the internet.
The fact that many of the bright computer science graduates from our country are moving abroad is an indication of the lack of rigour in our academics. The reasons behind this talent shift cannot be pinned to a single factor. What we can do instead is to offer coursework of incredible depth, and implement initiatives independent of what the policymakers have on their portfolio.
If we look at the top works being done in advanced fields within AI, we can see that there are many Indians involved and they have collaborated with top talents from the likes of Google or MIT. Surprisingly, most of these researchers also happen to have a background in IIT. The foundational education that is acquired at the undergraduate level is being leveraged outside India at a higher level.
There is no doubt that there is tremendous potential for the subcontinent to be the next big thing in the coming years. Since it is the premier institutes that are at the forefront of this transformation, it is crucial to be up to date in our pedagogical approaches.
For example, IIM Calcutta offers a one-year data science program that is unlike the online/hybrid programmes from ed-techs. IIM Calcutta, along with the Program Partner VCNow, brings the real classroom experience with live, face-to-face instructor-led sessions, delivered in a highly interactive, live two-way HD Video Conference-based virtual classroom platform across 30 cities. Those who cannot make it to the centers, can avail the direct to desktop option as well.
Whereas, IIT Madras faculty are offering affordable courses on data science through their platform ‘PadhAI.’ They are launching an online self-paced course on foundations of data science, which covers beginner-level concepts to mathematical and programming skills required for a data scientist.
This five-month course is affordably priced at ₹1,000 for students and faculty members. ‘PadhAI’ is the online school of One Fourth Labs, an IIT Madras incubated startup founded by Dr Mitesh Khapra and Dr Pratyush Kumar, assistant professors in the department of computer science and engineering, IIT Madras.
Including advanced domains in the curriculum as well as inculcating an ecosystem of partnerships with industries on a large scale is the only way forward.