Anomalyze this

Project and Program: Information Management, Database

Tags: Proceedings, SHARE Virtual Experience 2021

Anomaly detection is a trendy topic ; and it is something that can be quickly explored by ourselves whether it's for your weekly Db2 trends report ; monitoring or capacity planning. Db2 12 for z/OS provides a couple built-in functions that can be used for this purpose and for quick Exploratory Data Analysis ; enabling people to do more analytical work within their daily roles. On this presentation ; we will take a look at several ways to do Anomaly Detection by ourselves and have an overall view of what these alternatives offer ; tips and gotchas. We'll begin using pure Db2 SQL and then move to Data Science tools using Python Jupyter Notebooks. Objectives: The audience will learn what Anomaly Detection is and ways to approach this task. The audience will learn new built-in functions in Db2 12 for z/OS and Db2 LUW 11.1 to do Anomaly Detection using only SQL. This includes pros and cons of this approach. The audience will learn more ways in which Anomaly Detection is handled through a few Data Science Python libraries.

Back to Proceedings File Library