liveBook · Manning

Eigenvalue

As explained in Build AI Drug Discovery Pipelines

An eigenvalue is the scalar associated with an eigenvector that describes the factor by which a linear transformation scales that vector.

An eigenvalue measures how much variance a corresponding eigenvector captures when PCA analyzes the covariance matrix of standardized molecular data. Larger eigenvalues identify directions carrying more information. section 6.2.3

Use eigenvalues in PCA to:

  • Rank eigenvectors: the largest eigenvalue determines PC1, followed by PC2 and later components. section 6.2.3
  • Choose the retained dimensionality: apply an eigenvalue threshold, retain enough components for a target proportion of variance, or inspect a scree plot. section 6.2.3
  • Compress a molecular dataset by retaining the first k principal components and representing each sample in the new coordinate system. section 6.2.3
Figure 6.9 PCA defines a new coordinate system based on principal components, where the important dimensions are self-evident, and we can reduce the data dimensionality to k by retaining the first k PCs as ranked by the amount of variance they explain.

Visualize how PCA ranks new coordinate directions by the variance their eigenvalues explain, making the retained dimensions evident. section 6.2.3

How Build AI Drug Discovery Pipelines covers it

How deep it goes
Explains eigenvalues through covariance, eigenvectors, component ranking, and retention criteria, but stops short of deriving the computation.
Examples
Uses standardized molecular features from small-molecule binding data and compares eigenvalue, variance, and scree-plot criteria.
What it assumes
Assumes familiarity with features and variance while defining covariance, eigenvectors, principal components, and standardization in context.
What you can do after
Evaluate how many PCA components to retain using eigenvalue thresholds, explained-variance targets, or scree plots.
How it connects
Connects eigenvalues to PCA, eigenvectors, covariance matrices, explained variance, and dimensionality reduction.

Also explained in

Compare how 6 books explain eigenvalue →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Build AI Drug Discovery Pipelines. The section links point to the parts of the book it draws on.