Commoditization Is The Biggest Problem In Data Science Education: Prof. Raghunathan Rengasamy, IIT Madras

The buzz around data science has sent many youngsters and professionals on an upskill/reskilling spree. Prof. Raghunathan Rengasamy, the acting head of Robert Bosch Centre for Data Science and AI, IIT Madras, believes data science knowledge will soon become a necessity.

IIT Madras has been one of India’s prestigious universities offering numerous courses in data science, machine learning, and artificial intelligence in partnership with many edtech startups. For this week’s data science career interview, Analytics India Magazine spoke to Prof. Rengasamy to understand his views on the data science education market.

With more than 15 years of experience, Prof. Rengasamy is currently heading RBCDSAI-IIT Madras and teaching at the department of chemical engineering. He has co-authored a series of review articles on condition monitoring and fault detection and diagnosis. He has also been the recipient of the Young Engineer Award for the year 2000 by the Indian National Academy of Engineering (INAE) for outstanding engineers under the age of 32.

Of late, Rengaswamy has been working on engineering applications of artificial intelligence and computational microfluidics. His research work has also led to the formation of a startup, SysEng LLC, in the US, funded through an NSF STTR grant.

---

---

Edited Excerpts: 

AIM: What are your thoughts on the data science education market in India, especially compared to the global market?

Prof. Rengasamy: NASSCOM reports have indicated that 70% of the Indian workforce would need to reskill themselves in data science areas by 2025. The hype around data science, machine learning, and artificial intelligence being niche qualifications that could land individual lucrative pay packages have made youngsters and professionals wanting to upskill/reskill themselves. This has led to unprecedented growth in the number of players offering data science education in India. Universities and industries are also recognising the need to reduce the demand and supply gap in the domain and encourage students and professionals to upskill themselves to keep up. 

Currently, the data science education market in India is crowded. While there are several players in the data science training arena, the need for top-notch programs that combine theoretical and practical applications of data science continues to exist. Globally, the same scenario persists. The pandemic has also led to universities abroad open out their programs to students willing to study online. Furthermore, institutions offering MOOC courses worldwide have identified a new business model wherein the end-user pays only when they need certification. This has led to the availability of a surfeit of resources to the interested student.

AIM: How did COVID impact the data science education market?

Prof. Rengasamy: The COVID 19 pandemic has impacted our lives in many ways. Organisations, educational institutions, and individuals are all adapting to working from home. In these uncertain times, people are constantly worried about retrenchment. This fear and the fact that online learning and degree programs are gaining legitimacy are pushing people to seek more data science courses.

AIM: What are the key challenges a university faces in instrumenting a data science and analytics course?

Prof. Rengasamy: Data science as a field has grown at an unprecedented rate. As a result, universities are hurrying to put together data science programs as the demand is dragging the supply along. 

Hundreds of universities have traditional engineering programs, such as chemical engineering, and these institutions act as breeding grounds for future faculty members. Unfortunately, not many such programs exist in the field of data science. As a result, there is a shortage of well-trained faculty. 

Further, the job market is so lucrative that it becomes difficult for universities to lure the best to come and teach. This is the biggest immediate challenge that needs to be addressed. The second challenge is identifying the correct mix of fundamentals and practical aspects that need to be taught. This is a general challenge for all fields, given the frenetic pace of development nowadays. This is more so in data science, where the field changes at an extreme pace.

AIM: What are the significant obstacles data and analytics education is facing in India?

Prof. Rengasamy: The recent Global Skills Index report 2020, which was released by Coursera, ranks India at number 51 and classifies Indian students as ‘lagging’ in trending skills such as data science, artificial intelligence, deep learning, and comprehensive understanding of topics such as linear algebra. This ranking may be attributed to interest in certification and finding jobs taking precedence over learning. 

The biggest problem we face today is the commoditization of education. Individuals and corporations alike would like quality courses to be offered by the best faculty at the lowest price. Of course, there are very talented individuals whose quest for knowledge is insatiable. It is critical that we goad students towards understanding key concepts, enjoy the learning process and encourage creativity. 

AIM: How governments and corporations can play a role in encouraging more students to select data science subjects?

Prof. Rengasamy: Many jobs in governmental organisations can benefit from the use of data science. The government can make this a part of their mandate, which will encourage students to pursue data science subjects. On the other hand, corporations are already pushing their employees to become ‘data-savvy and asking them to upskill and reskill. This is one of the reasons for the popularity of this field. 

Corporations can also open up a lot more internship opportunities to encourage students to dive into this dynamic field.

AIM: What is your projection for AI and big data analytics education growth in the future?

Prof. Rengasamy: The training market for emerging fields such as machine learning and AI is poised to grow at a double-digit rate. These claims are substantiated by several findings.

Talent demand and supply — AI and Big Data Analytics report by NASSCOM indicates the growing demand for data scientists. As per the report, the total demand for big data and AI talent is estimated to be approximately 800,000 in the next couple of years. The demand-supply gap is expected to reach around 140,000. 

On the other hand, according to the World Economic Forum report, AI may lead to a net increase of 58 million jobs globally. The forum also estimates that around 40% of workers will require reskilling in six months or less.

Another report by Internshala Trainings highlighted ethical hacking and machine learning as the major skills that Indian students learned in 2020 to prepare themselves for the career opportunities of 2021. Further, a recent HBR article stated that millions of workers would need to be retrained or reskilled due to AI penetration over the next three years.

AIM: How can an industry partnership add value to data science courses offered in the market?

Prof. Rengasamy: Fundamentally, the skill gap between what the industry needs and the talent that job seekers possess can be attributed to the gap between what is taught at the university and the actual challenges in the industry. The concept of internships has evolved to bridge the skill gap. As with other domains such as engineering and accounting, industry partnerships will add immense value to data science courses. 

Learning data science techniques and algorithms alone might not be enough. Developing critical thinking to formulate problems, understand data requirements, and propose solutions that respect practical considerations are important skills. Exposure to real-life problems will help students develop these skills. 

AIM: What’s your advice for aspiring data scientists?

Prof. Rengasamy: Availability of data has led to the widespread application of data science techniques in all fields. Knowledge of data science basics will become a necessity soon, pretty much like how the job of an office administrator has moved away from performing simple tasks to including knowledge of computer software. However, the expertise one needs or the expertise that one can attain will vary. It is important to understand that someone who uses NLP need not be able to develop the algorithms themselves. Effectively, depending on the level of expertise one can gain, a wide range of opportunities are available.

Open-source tools have led to the democratisation of knowledge. Aspiring data scientists need to focus on honing their critical thinking skills and developing a strong understanding of fundamental topics relevant to data science, such as linear algebra, optimisation, and statistics.

AIM: What does the future look like for individuals pursuing a data science career?

Prof. Rengasamy: Today organisations have significant amounts of data and seek data scientists to derive insights from these data. We are more equipped than ever before to derive insights from data. The transdisciplinary enquiry has become the need of the hour. 

The field requires people from all domains and is not limited to computer science graduates or engineering graduates. Healthcare, fintech, ecommerce, marketing and economics have all utilised available technology to make great strides. Like with any other profession, the level of experience, domain expertise, and associated skills will determine an individual’s position in the data science pyramid.

AIM: Considering there are many online courses and MOOCs available for data science enthusiasts, how do you think a professional degree makes a difference in their career? 

Prof. Rengasamy: A good professional degree program provides structure to one’s learning. It is clear that all the resources needed for someone to learn data science are all freely available online. However, learning requires the availability of resources and is also dependent on the learner’s motivation, perseverance, and, of course, their innate ability. A small fraction of learners can simply learn from online resources. However, other learners would require considerable handholding and guidance.  

Incidentally, a well-structured basket of courses is what an online degree provides. This aspect is not only relevant to data science but also to all fields. For example, one would be able to find an online course mirroring every course in a degree program in chemical engineering. However, one does not pose the same question about traditional engineering disciplines. This is because we find it hard to imagine that one can claim to have mastered aeronautical engineering through only doing MOOCs. 

This difference is because we perceive a limited skill-set as being enough for someone to claim themselves as a data scientist or engineer. A lot of this could also be attributed to the commoditization and packaging of general-purpose algorithms that allow practitioners to do a lot more with extremely limited understanding. In essence, the choice of a professional degree or a set of MOOCs depends on where in the data science pyramid one wants to be.