AI, machine learning and deep learning · Intermediate

WHAT ARE EIGENVECTORS AND EIGENVALUES?

Written by SignalsLab

Early edition: these lessons are being reviewed.

Content

Summary

Eigenvectors are vectors that define the specific directions along which a linear transformation acts by flipping, compressing, or stretching. Eigenvalues represent the strength of this transformation, quantifying the factor by which the eigenvector is scaled in that direction. In data analysis, these concepts are typically applied to correlation or covariance matrices to understand underlying structures.

Intuition

Think of a linear transformation as a force that distorts space. Most vectors will change both direction and magnitude when this force is applied. However, eigenvectors are special because they remain on their own line; they only get stretched, compressed, or flipped, but they do not rotate. The eigenvalue tells you exactly how much stretching or compression happens along that specific eigenvector direction. If you are analyzing data, you often look at correlation or covariance matrices to find these special directions and their associated strengths.

Derivation

  1. Identify a linear transformation acting on a vector space.
  2. Observe that most vectors change direction and magnitude under this transformation.
  3. Identify specific vectors that do not change direction, only magnitude or sign.
  4. Define these invariant direction vectors as eigenvectors.
  5. Quantify the scaling factor applied to each eigenvector as the eigenvalue.
  6. Apply this framework to correlation or covariance matrices in data analysis contexts.

Worked example

Problem

Consider a linear transformation that stretches space by a factor of 3 along the x-axis and flips it by a factor of -1 along the y-axis. Identify the eigenvectors and eigenvalues.

  1. The transformation acts along the x-axis and y-axis directions.
  2. Along the x-axis, the space is stretched by 3, so the x-axis direction is an eigenvector with eigenvalue 3.
  3. Along the y-axis, the space is flipped (stretched by -1), so the y-axis direction is an eigenvector with eigenvalue -1.
  4. These directions are the eigenvectors, and 3 and -1 are the corresponding eigenvalues.

Answer Eigenvectors are the x-axis and y-axis directions; eigenvalues are 3 and -1 respectively.

Common mistakes

  • Confusing eigenvectors with general vectors that change direction under transformation.
  • Misinterpreting eigenvalues as angles rather than scaling factors.
  • Assuming eigenvectors are only relevant in abstract math rather than data analysis contexts like covariance matrices.

Animation

Animation coming soon

The animated version of this lesson is being prepared for the web. In the Signals AI app, lessons are drawn step by step.

Code

This is a text lesson: it has no animation code yet.

ccBYNC© 2026 SignalsLab. Lesson text shared as is under CC BY-NC 4.0.