A Two-Step Process for Detecting Fraud using Oracle Machine Learning, APEX & Oracle Analytics Cloud

Oracle Machine Learning comes free inside every Oracle Database EE, SE2 DBaaS and the Autonmous Databases. Oracle Machine Learning extends Oracle Database(s) and enables users to build “AI” applications and analytic dashboards. OML delivers 30+ powerful in-database machine learning algorithms, automated ML functionality via SQL APIs and integration with open source Python* and R.

First, we use an unsupervised Oracle Machine Learning algorithm (1-Class Support Vector Machine) in the Oracle Autonomous Database (ADW) to “flag” anomalous insurance claims.

We use a simple Oracle APEX application for the claims investigators to focus on the most suspicious claims using their expertise and knowledge to enter their fraud/no fraud decisions.

We then build Oracle Machine Learning supervised learning classification models in ADW on the "labeled" data (FraudFound = Yes/No).

Finally, we build interactive APEX and Oracle Analytics Cloud dashboards on our fraudulent claims discoveries and then automate the data-driven process

Overview: 00:00
Unsupervised Learning with ADW + OML: 12:56
Labeling Claims as Fraud/NoFraud using Oracle APEX: 17:58
Supervised Learning using ADW + OML: 21:50
Reviewing Results using Oracle APEX: 27:49
Interactive Data Analyis of OML Predictions and Insights using Oracle Analytics Cloud: 31:35
Resources/Links: 37:01