Thanks. I'd love some ELI5s on this.

My last gig had some newly minted CS graduates who proposed complimenting our monitoring with some anomaly detection.

Having done a bit of both data mining and optimization, a lifetime ago, I could kinda follow the big data and machine learning stuff the kids were doing.

Whereas anomaly detection and fault prediction seem magical to me.