Mathematics for AI and Machine Learning

Foundations for modern AI and machine learning

LDL^T Decomposition

Chapter 9 Linear algebra

Chapter 9: Matrix Decompositions and Beyond — LDL Decomp

High-resolution PNG
LDL^T Decomposition — high-resolution mind-map icon

From the book

Chapter 9: Matrix Decompositions and Beyond. In the chapter mind map this icon labels LDL Decomp: $A = LDL^\top$. The discussion below is excerpted and lightly edited from § Theorem: LDL Decomposition in Mathematics for AI and Machine Learning.

For a __symmetric__ matrix $A \in \mathbb{R}^{N \times N}$ that admits LU decomposition, there exists a decomposition

where $L_1 \in \mathbb{R}^{N \times N}$ is a lower triangular matrix with unit diagonal, and $D \in \mathbb{R}^{N \times N}$ is a diagonal matrix. We use the notation $L_1$ (with subscript 1) to emphasize that it has unit diagonal, distinguishing it from $L$ used in other decompositions (e.g., Cholesky) where the diagonal may not be 1.

This is the symmetric version of LDU decomposition, where the symmetry of $A$ implies $U = L_1^\top$.

What this drawing shows

What you see. Shows a symmetric matrix factored into lower-triangular, diagonal, and transpose factors without square roots.

In the mind map. Chapter 9 — LDL Decomp. See From the book above for definitions, figures, and worked examples.

Where to read next

Open Chapter 9 companion →

Read the full definitions, figures, and worked examples in Chapter 9: Matrix Decompositions and Beyond — see the mind-map node LDL Decomp.