What got you into Data Science/AI?
I got into Data Science/AI because I find applying mathematical and scientific reasoning to solve complex real-world problems that are of high economic value very satisfying. Business decisions – both involved and simple – are constantly being made in the business world. Should we make these decisions based on our intuition and gut feel or scientific reasoning and data analysis? Complex decisions encompass too many variables for the human mind to comprehend and analyze. The amount of available data and the ability to process it with machines is growing exponentially, whereas human brain processing capacity is limited. Data Science/AI allows us to model complex real-life situations and provides optimal decision-making capability in the context of global business objectives.
What is most exciting for you in Data Science/AI?
Einstein is quoted as having said that if he had one hour to solve the problem he would spend fifty-five minutes thinking about or defining the problem and only five minutes in thinking about or finding the solution. In other words, before jumping right into solving a problem, we should step back and invest time and effort to improve our understanding of it. And that is what Data Science/AI enables us to do.
Data science has the power to improve our decision making and problem-solving ability by helping us to explore new ideas and acquire new knowledge through the collection, analysis and interpretation of data. It enables us to capture domain characteristics and allows us to solve seemingly intractable problems without sacrificing the global optimality as opposed to relying on “general purpose” techniques and heuristics.
What do you think is the biggest challenge faced by Data Scientist today?
The theory of computational complexity has demonstrated that general purpose problem solving paradigms are not going to allow us to get optimal solutions to complex (NP-complete) problems efficiently. The conventional wisdom of designing good “heuristics” is flawed in the absence of human experts. Solving such intractable problems remains a big challenge. We envision a future where a team of software agents that rely on AI with big-data and machine-learning techniques will ultimately enable us to find practical models to solve such complex problems.
Today, a modeler is faced with the very difficult choice of determining what the best model to develop is. If the model is made too complex it may not be possible to complete the model with the time, knowledge and supporting data available. If the model is too simple and the results obtained may not be sufficiently accurate. One of the most difficult issues in business decision optimization modeling is determining the right model that needs to be developed to capture the real world. The job of the modeler is to understand the real world of business optimization where the outcome is a function of myriads of variables and the market is intrinsically somewhat indeterminate, and there is uncertainty associated with any forecast.
On the surface it might appear that the answer is to build the model that contains as much detail as possible. After all, this model will be the closest to the real world and so surely the most accurate. This might be true if we had complete knowledge of the real world and a very large amount of time available to develop and run the model. But, we only have limited knowledge of the real world or the real world is intrinsically uncertain, and we have limited time to find a solution as the problem has to be solved in real time. Indeed, we rarely have the luxury of living in a deterministic world where the future can be forecast with no uncertainty and we have both complete knowledge and enough time to decide.
What do you think is the biggest obstacle for Data Science/AI adoption?
So, what’s holding many firms back? There are several factors, but two of the biggest impediments to greater adoption are access and knowledge.
The Data Science/AI practitioner should expect to meet a normal amount of resistance to change to the new concept and should be prepared to deal with it in a constructive manner. As with any new process, the people who will resist the innovation are those who are concerned that their existing skills and corporate status may be devalued by the new procedures, and/or those who find that the innovation creates new obstacles to the achievement of their existing goals. These are the people who are most likely to feel threatened by the potential devaluation of their skills and may fear that the new technology will supplant their expertise and render them unimportant or even superfluous. It is also common for these people to fear that they may not be smart enough or well educated enough to manage this Data Science/AI “sophisticated” technology. Such fear is generally groundless and can be overcome by coaching as explained later.
Which is your favorite session at ODSC India that you looking forward to attend?
I am interested in all aspects of Data Science, Automation of Intelligence, Deep Learning and Machine Learning. I am looking forward to attending most, if not all sessions at ODSC India.
You are traveling all the way from US to Bangalore. What got you interested to present at this conference?
This conference gives me an opportunity to connect with the most innovative people in the field of data science in the country I was born in and feel proud that India has the most number of data analytics jobs after US!
Any personal message/remarks you want to share with the DS/AI community in India?
I would like to share two quotes I love. The first one is a quote from my favorite mathematician & scientist Albert Einstein – “Imagination is more important than knowledge. For knowledge is limited, whereas imagination embraces the entire world, stimulating progress, giving birth to evolution.” The second quote – I don’t know who it is attributed to – that has helped me all my life is “The quickest way to acquire self-confidence is to do exactly what you are afraid to do.”