liveBook · Manning

Underrepresentation

Underrepresentation is the insufficient presence of a population, class, structure, or condition in data relative to a relevant population or distribution.

How 2 Manning books explain it

Build AI Drug Discovery Pipelines

Underrepresentation is insufficient inclusion of a population, class, structure, or condition in a dataset relative to its relevant real-world distribution.

How it is used
Frames underrepresentation as a data-quality limitation that restricts drug-discovery model applicability.
How deep it goes
Compares dataset proportions with real-world distributions and traces effects on accuracy and generalization.
Examples
Uses membrane proteins, macrocycles, larger molecules, and underrepresented patient populations as concrete cases.
What you can do after
Evaluate how skewed training data can limit prediction accuracy and generalization across targets and patients.

Read how Build AI Drug Discovery Pipelines explains underrepresentation →

Machines that Think

Underrepresentation is the insufficient presence of a group or its examples in data, systems, or decision processes relative to the relevant population.

How it is used
Frames absence from data as a source of bias in language technologies and automated decision systems.
How deep it goes
Traces how sparse representation can shape model responses and system outcomes through several concrete cases.
Examples
Uses healthcare-cost prediction, speech recognition, gender classification, dialect moderation, and predictive policing examples.
What you can do after
Evaluate how sparse representation can distort outputs and contribute to unequal treatment across groups.

Read how Machines that Think explains underrepresentation →

Related concepts

Questions about underrepresentation

What is underrepresentation?

Underrepresentation is the insufficient presence of a population, class, structure, or condition in data relative to a relevant population or distribution.

Which Manning books explain underrepresentation?

Build AI Drug Discovery Pipelines and Machines that Think.

What is underrepresentation a kind of?

Underrepresentation is a kind of representational harm. Representational harm is damage caused when language or information technologies negatively, inaccurately, or inadequately portray, erase, or distort individuals or groups.

These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.