The attached article from TechRepublic asks the reader which they would trust more, predictive Analytics, or their gut. I would say this is a lot like police work. Once you’ve walked a beat long enough, you get to know the neighborhood. You get a feel for the people, the spots with the most questionable activity, and a sense of times when questionable behavior might occur. If you are, let’s just say, in the IT field, years of experi NCR can help you learn applications and networks in a similar fashion.
The same thing could be said for financial markets or other industries as well. Does this intimate knowledge trump a computer’s ability to crunch numbers based on real data? No. The machine learning can also remove some of the emotion or empathy that might get in the way of a cop performing their duty. My answer to this question is I trust both, or a marriage of the two.
I want the predictive models, but I want to see them and apply my own logic and understanding to that model. This can be influence both before and after the models are calculated. Augmenting our knowledge and skills with predictive analytics based on true historical data is the best way to achieve the highest success rate. That is my two cents.