This reference is partly in my sweet spot of coverage, and it is partly beyond it pointing toward enterprise. And no, it is not predictive maintenance, by the way.
This is a post on the SAP blog about 10 Myths of Predictive Analytics. It is quite accurate about many of the “higher-end” software applications.
It’s not easy to do right. It requires insight as well as pure number crunching to obtain maximum benefit. It is iterative; that is, just like advanced process control you don’t run it once and declare it done. Maximum benefit means iterating the data and process.
It is a good read.
