Feature

A feature is an input property or representation used by a model when producing an output.

Plan for this page

Understand one concept in plain language and connect it to nearby ideas.

You will leave with: Features can be directly measured, engineered from raw data, or learned internally. Their meaning ranges from an interpretable value such as temperature to a dimension in a representation that has no simple human label.
Time: 3–8 min
Before you begin: No prerequisite. Keep one real example in mind.
Do this now: Read the explanation, then compare the example with the common misunderstanding.

Prepared next stepTrace data into model behavior

FeatureA feature is an input property or representation used by a model when producing an output.

Model
Dataset
Label
Embedding

Plain explanation

Features can be directly measured, engineered from raw data, or learned internally. Their meaning ranges from an interpretable value such as temperature to a dimension in a representation that has no simple human label.

Example

A loan-risk model might receive income, existing debt, and payment history as features, subject to legal and fairness constraints.

Why it matters

Feature choices determine which information the model can use and may introduce leakage, proxies, privacy exposure, or missing context. More features are not automatically better.

Common misunderstanding

Misconception: every feature is a harmless, objective fact. Correction: features can be noisy, sensitive, derived, correlated with protected traits, or unavailable in real use.

Related concepts

Follow relationships, not only the alphabet

ModelA model is a learned or designed representation that maps inputs to outputs for a defined task or family of tasks.DatasetA dataset is an organized collection of examples or records used for training, evaluation, analysis, or system operation.LabelA label is a target value or category attached to an example for training or evaluating a supervised model.EmbeddingAn embedding is a numerical vector that represents an item so useful relationships can be measured or learned.

Trace data into model behavior

Sources and scope

What supports this object

Machine Learning GlossaryGoogle for Developers

Defines features in machine learning; modern models may learn internal representations rather than rely only on manually selected features.

Open source

Loading page content...

…