Data Science is the latest buzzword in the industry and is looked upon as the next big thing in the industry. Its growing popularity can be identified from a recent study by Glassdoor that identified Data Scientists as the top job in the USA. And the scenario is no different in India. An increase in the amount of data generated by industries is one of reasons why data science has become so popular over the years.
S Vaitheeswaran, Managing Director and CEO, Manipal Global Education, during a talk at Cypher 2017 quickly confronts that during a meeting with fresh grads in one of the Manipal Universities, there was repeatedly a mention about artificial intelligence, machine learning and a willingness to make a career in data science.But is data science and related technologies such as AI being explored to the extent that it should be?
A difference of opinions in these areas, especially if AI is good for people or not, has been quite prominent—and it was recently witnessed in the form of famous Musk and Zuckerberg fight. Not just AI, but an increasing accessibility to personal and sensitive data such as that from Aadhar has invited debates of privacy breach and other risks. Vaitheeswaran remarks that in today’s world others may know you better than you know yourself—thanks to the ease of collecting data and accessing it.
Data Science—a disrupting area with challenges
“Any disruption that happens brings opportunities. We saw it thirty years back in the IT world and now a disruption in the area of data science, analytics is enabling companies, young startups, government and professionals to leverage it”, said Vaitheeswaran. He also noted that amidst the gaining popularity that it offers, it is important to make this subject more encompassing, pervasive and avoid hubris.
There is no doubt that data science has become a hugely adopted area in most companies, but most of them jargonize the area way too much, to an extent of making it look a little hostile. “Subjects like data science and analytics should become more pervasive and we must open it out to everybody in the system. People must not find it difficult to understand the subject”, said Vaitheeswaran. After all, it is important to note how data science can impact people in everyday life. At the end of the day, data science is not just about spreadsheets and algorithms but in understanding how everything in everyday life can relate to data science.
Not just technology and business, but people matter too
Given the adoption of any technology, at the end of the day it is important to find solution for various problems. “You are going to miss the big picture if you don’t relate to the problem”, said Vaitheeswaran. He further exemplified that Manipal hospital extensively makes use of Watson in everyday life but do the makers of Watson really understand the domain for which they are solving the problem? Have they ever been to the operation theatre? “They make solutions but may not come in terms with the real time implementation of the products that they are offering for various industries”, he said. This remains a major challenge.
The other thing is how this space can impact the society. “It may be great to work for the companies and people, but in a country like ours, it is important to understand how big data, analytics and similar technologies can impact our social ecosystem”. There are companies coming in this space using predictive analytics, to manage the crowd in Kumb Mela, for instance. They are using drones and creating backend database to solve problems such as stampedes.
It is time that more companies come into this space with an intention of serving the community, to make this industry larger by using complex tools. “The appetite to learn is good in the country and therefore it should be used to explore larger areas”, he said. Though a lot of growth is happening in this area, such as use of deep learning to understand the ocean, using predictive analytics to reduce the effects of natural disasters, people should try working with social organizations and NGOs more.
On a concluding note he said that dashboard for soft metrics are certainly more difficult than dashboards for operating metrics that we all are comfortable with. “For me this missing D in data science could be domain, direction or the demographic dividend. Rest it’s a responsibility for all of us to identify the missing D”.
Try deep learning using MATLAB