Bank Fraud: Using Technology to Combat Losses

CHAPTER 9

The End: It Is Really the Beginning!

History has witnessed many “Golden Age” periods. We are now entering the golden age of predictive analytics. The last several decades have been spent accumulating data meticulously. The race to make sense of this data is on. Being able to develop insight into the data gathered—from a historical context as well as a predictive context—is becoming the primary driver of many businesses.

As a statistician, I feel very gratified to live in this era. At the beginning of my career, I worked as a biostatistician and still remember vividly how lucky and happy I felt when a surgical experiment was conducted on 30 rats as opposed to 15 rats. At least we could pretend the law of large numbers was being used. Certain areas of biostatistics still have to rely on very little data to draw some key conclusions. It is the nature of clinical research, especially when the researchers are trying to establish a procedure as valid. It can be prohibitively expensive to conduct an experiment on a large number of specimens to establish validity. Once an idea is validated, a much broader set of subjects can be used for the research. In areas of research, the issue of having to use such limited data probably cannot be totally eliminated. However, behavioral modeling can be based on a lot more data today than was possible even ten years ago.

The capturing and storing of the data have become really inexpensive. Terabytes don't have to be “terror” bytes anymore. ...