Glossary¶

In this first module, we focus only on:

• Detect deviations in a static dataset using clearly defined domain rules.

Anomaly¶

An anomaly is a value that:

• Falls outside the expected or reasonable range for a given context, and/or
• Violates the domain rules defined.

Important:

An anomaly is not universal. It depends on assumptions, context, and domain knowledge.

Sometimes we have real-world information about reasonable thresholds for:

• Age in years for humans
• Age in years for human children
• Age in years for dogs
• Age in years for cats
• Age in years for tortoises
• Height in inches for humans
• Height in inches for human children
• Height in inches for dogs
• Height in inches for cats
• Height in inches for tortoises

Understanding Comes First¶

Continuous Intelligence does not begin with complex models.

It begins with clarity and experience:

• What do we expect?
• What violates that expectation?
• How do we detect and document it?