Feature Space

Feature #

A feature is an individual measurable property or characteristic of a data point used as input to a machine learning model.

Each feature corresponds to one dimension.

\[ x_i \in \mathbb{R} \]

A data point with ( d ) features is represented as:

\[ \mathbf{x} = \begin{bmatrix} x_1 \\ x_2 \\ \vdots \\ x_d \end{bmatrix} \]

Geometric Intuition #

A feature is like one axis in a coordinate system.

  • One feature → line
  • Two features → plane
  • Three features → 3D space

Each new feature adds a new dimension.

Why Features Matter #

Models learn patterns in features, not raw objects.

Good features:

  • Capture relevant information
  • Make patterns easier to detect

Poor features:

  • Add noise
  • Increase dimensionality unnecessarily

Feature Space #

Definition #

A feature space is the vector space formed by all features, where each data point is represented as a vector.

\[ \mathbf{x} \in \mathbb{R}^d \]

Here, ( d ) is the number of features.

Geometric Intuition #

A feature space is a coordinate system:

  • 1 feature → line
  • 2 features → plane
  • 3 features → 3D space
  • Many features → high-dimensional space

Each data point becomes a point (or vector) in this space.

Relation to Linear Algebra #

Feature spaces are:

  • Vector spaces
  • Often high-dimensional
  • Manipulated using matrices and vectors

Operations such as projection, rotation, and dimensionality reduction occur directly in feature space.


Home | Vector Spaces