Machine learning is essential in Oracle 19c's Automatic Indexing, enabling the automatic creation, management, and optimization of indexes based on workload patterns.
Observing Patterns: The system watches which SQL queries are run most frequently and suggests or creates new indexes for the columns most often used in those queries.
Adjusting Over Time: The system can remove or change indexes that are no longer used to keep things efficient.
Machine learning in automatic indexing significantly improves database performance by intelligently managing indexes according to real-time data and evolving patterns.