Saturday, May 18, 2013

SMALL DATA BEFORE BIG DATA


How did humankind evolve all the way to the Big Data Revolution?



At first there was the Age of statistical sampling  and extrapolation of Small Data…



… which although with some degree of bias, could give considerable support to essential predictive decision-making…


…but only now we have the luxury and the precision of interconnected massive data-sets that allows us to make accurate analysis and selections…


…although again with some degree of risks when trying to orient in the Big Data ocean…

In short, Big Data is a Revolution and a big step forward as it saved us from analyzing random samples full of individually biased errors by drowning them under huge data-sets. However, distinguishing useful data from useless data, building quick analytical processes and ensuring customer privacy represent the next challenges for Big Data statisticians. It is also important not to enthusiastically rely on data as something that never fails as it could turn into some significant unfair decisions such as refusing insurance or credit to a potential customer only because an algorithm says so.

1 comment: