Big Data, Predictive Modelling & Economics

Introduction:

What is data useful for? What are some of the techniques and Big Data technologies that we hear about every day? Measuring information, extrapolating them onto sheets, and finally predicting a surmised model, where does all this leads us to? Well when we are working with data we are essentially handling certain quantified information that would help in prediction and forecasting. These predictions and forecasting go a long way in removing uncertainties.

An Uncertain Tale:

Lucky John wanted to sit for an engineering based MCQ test. He wanted to clear this test as this would allow him to gain admission into his desired college. For this he practiced mocks regularly. He realized that if he got lucky and scored really well on the test day that would be enough to see him through but he also realized that his result would be purely based on luck and he wouldn’t be lucky all the time. However, if he consistently scored well, in the mocks, there was a much higher certainty that he would pull off a good score in the final exam.

This is essentially what data does. Big Data and Predictive Modelling increases certainty or reduces uncertainty.

When you are using Big Data:

• You are essentially fitting patterns.

• You are correlating things to a curve.

• You are predicting the likelihood of an event.

• You are reducing the likelihood of an adverse circumstance.

Economic Models Do Not Hold True In an Imperfect Economy:

We have often studied in Economics how in a perfect market and perfect competition at the equilibrium point where the demand curves meet, buyers can buy all they want and sellers can sell all that they can. However, certain discrepancies were found in the model which flawed the conclusion. Economists preferred to call them externalities. Nonetheless, at present we know that branding, advertising, subjective whims and biases play a much more profound role than was earlier thought.

Ultimately, the industries that employ quantifiable Big Data techniques stand a much better chance than their competitors for thriving in the market, due to a much more accurately predicted world.

Fault-free Model in an Unpredictable World Is Non-Existential:

Obviously there is no model that is absolutely free from flaw. We live in a skewed world filled with unpredictability and as such discrepancies do arise.

For example, when a car swerves and takes a desperate left turn, you can predict that an accident is about to take place the second before it happens. On the other hand, although cigarette smoking and cancer both cuts your biological life, still it would be difficult to predict which of the two individuals would die first or who would live and who would die in say, the next ten years.

However, you can predict with much certainty that both of these individuals would have led longer lives had they been free of the disease and the habit and that is precisely where Big Data techniques are so powerful.