liveBook · Manning

Linearly separable

As explained in Timeless Algorithms, The Foundational Ideas

Linear separation is the property that classes can be divided by a linear decision boundary in a feature space.

Two classes are linearly separable when a linear decision boundary divides their training observations without error. In one dimension, that boundary is a threshold. section 8.1.1

Separability establishes that a perfect boundary exists; it does not choose among multiple boundaries. A maximum-margin SVM selects the boundary farthest from the nearest observation in each class. section 8.1.1 section 8.1

Safe and unsafe temperatures

Plot manufacturing temperatures on a number line and label readings safe or unsafe. A threshold between the classes separates the observations when none fall on the wrong side. section 8.1.1

Figure 8.1 Temperature readings appear along a single horizontal axis, increasing from left to right. Circles represent safe conditions, and squares represent unsafe conditions. The vertical dashed line is the decision boundary. The nearest observation on each side is a support vector, and the distance between the boundary and these points defines the margin on either side. Positioning the boundary midway between the support vectors produces the maximum-margin separator.

The figure shows the threshold, nearest observations, and margin in the temperature example. section 8.1.1

How Timeless Algorithms, The Foundational Ideas covers it

How deep it goes
Defines error-free separation in one dimension and previews hyperplanes, margins, and support vectors before algebraic optimization.
Examples
Uses safe and unsafe manufacturing temperatures on a number line, then extends the synthetic classification example into two dimensions.
What you can do after
Understand when training observations admit an error-free boundary and why maximum-margin SVMs prefer one separator over another.
How it connects
Connects linear separation to decision boundaries, support vector machines, separating hyperplanes, margins, support vectors, and generalization.

Also explained in

Compare how 4 books explain linearly separable →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Timeless Algorithms, The Foundational Ideas. The section links point to the parts of the book it draws on.