This article is Talk 05 of Honest Data Science Talks series. Each talk of around 500 words length, is personal thoughts and deliberations on the domain.
Everything we learn has to be put to use someday and that has got to be the essence of education. Or they must be the driving factor to put together and solve a problem someday. They are like the tools in the tool-bag. Not all are needed every time. Some might have everyday use, some might just be for assistance, some might be emergency need only but it’s good to have all of them in the bag.
The goal of education is also to prepare one for real-life hurdles. Apart from routine 9 to 5, life presents real challenges. Even though one has IQ more than desired and if one cannot aid in decision making, it serves no purpose. Maybe the IQ was never exposed to real learning.
The question is not about having a better IQ, but about how to put that to use on the right platform.
Let’s take a hypothetical scenario that there is a new virus that is out there affecting thousands and there is an approximate pattern noticed. The growth is definitely exponential. There is sufficient data available on how it is spreading along with the demography data. What’s the use of data science if it cannot be used to comprehend the beneficial decisions? What’s the use of data scientists if they cannot plot the graph of when it would stabilise?
Where is the model that predicts fruit disease? The model that predicts leaf disease? The model that predicts cancer? The model that predicts [put-anything-omg-here]? The one that had an accuracy of 99%? Why isn’t anyone running them on real data? Why does the world still lack when it comes to real-life problem handling? I once attended a presentation where the team presented a model to evaluate the quality of the question paper. My concern was what’s the use? What’s the use of evaluating after the scores are in and students have already given the exam? What’s the use if it cannot be further used to devise the future question papers? It will be yet another thesis or paper on the shelf – that’s all.
I have read several articles lately about how math is no more required for coding. Of course, it is bread and butter of coding. If someone has given you matrix operations as API, you were hindered from learning the insights from the domain. If you don’t understand the Catalan series, if you don’t know the use of Fermat’s theorem, if you don’t understand Chomsky hierarchy, If you cannot apply Prisoner’s Dilemma, if you don’t understand Granovetter model, and the list follows, we got to question our-self and our learning. We aren’t preparing ourselves for the future. Data science needs a long-lasting self-learning future vision. The science is real when and if it can anticipate future challenges.
One of the recent business models is to label that the product can heal the [trending-virus-name] and share it on social media. There is one class of elite educated people who hoard unnecessary things and another class of people who are not aware and of what is happening around. The community needs to know that the solution is in the balance and not in the imbalance. To say, the educated community is creating more problems than uneducated.
The question is when we claim to model a science, do we have the right models to do that? Do we really have models purely driven by the data or are they purely statistical borrowings? We have several problems at hand occurring on the daily basics. We definitely need data science to find a truer meaning to it.