Large Scale Eigenvalue Problems : Single Vector Iteration
Single-vector iteration, also known as the power method, is one of the simplest ways to approximate the dominant eigenvalue and eigenvector of a matrix.
For a matrix \(A\), the basic idea is:
- \(A \in \mathbb{R}^{n \times n}\), with \(n\) distinct real eigenvalues
- And the corresponding eigenvectors are:
- We can express a random starting vector \(x_0\) as a linear combination of eigenvectors:
- By definition of an eigenpair:
- Repeated multiplication by \(A\) gives:
- After repeated multiplication, the dominant component remains:
- Since $x_{p+1}$ is only approximately a scalar multiple of $x_p$, we estimate the scalar by projecting one vector onto the other:
- To prevent underflow or overflow, we normalize the eigenvector estimate:
The explanation above was taken from Power method of eigenvalues - Lesson 30 - Numerical Methods for Engineers.
The animation below illustrates the iteration process.
WOLF-I