Machine Learning scoring and Rule engines in z/OS

Making better Decisions with Rules and Machine Learning on z/OS
This article extends the topic implementation discussed in the Redpaper of IBM’s Redbooks
Machine Learning with Business Rules on IBM Z: Acting on your insights

Article – Making better Decisions with Rules and Machine Learning on z/OS

Summary:
Operationalizing Machine learning in realtime by using Machine Learning scoring with Business Rules. This article describes how the two technologies can be used together in real time.

The Redpaper looked at both technologies and described how the two can be used together. This Development article extends this and describes a real time solution.

The ‘Machine Learning scoring engine’ runs as a feature in the Liberty profile from IBM and the ‘Operational Decision Manager WOLA Rule engine’ can also be hosted in a Liberty server. WOLA stands for WebSphere Optimised Local Adaptor and acts as a high performant in-memory pipe between client applications and a Liberty server.

If we can configure Liberty to host both engines in the same Liberty then we could potentially have the rule engine call the scoring service direct to make better decisions and in real time.