# 31. Eigenvectors of Circulant Matrices: Fourier Matrix | Summary and Q&A

33.4K views
May 16, 2019
by
MIT OpenCourseWare
31. Eigenvectors of Circulant Matrices: Fourier Matrix

## TL;DR

Circulant matrices, which have a cyclic structure, are closely connected to the discrete Fourier transform and are important in applications such as machine learning and image processing.

## Install to Summarize YouTube Videos and Get Transcripts

### Q: What is the significance of circulant matrices in machine learning and image processing?

Circulant matrices are used to reduce the dimensionality of data in machine learning and image processing. By using a circulant matrix, it is possible to describe an image using a smaller feature vector, which simplifies computations and makes the processing more efficient.

### Q: How are circulant matrices related to the discrete Fourier transform?

Circulant matrices are connected to the discrete Fourier transform because they share similar properties. Like the discrete Fourier transform, circulant matrices have a cyclic structure and can be used to compute circular convolutions. This makes circulant matrices useful in applications where shifting and convolution operations are important, such as signal processing and image analysis.

### Q: How do circulant matrices differ from other types of matrices?

Circulant matrices have a special structure where each row or column is a cyclic shift of the previous row or column. This is different from general matrices, which can have arbitrary entries in each position. Circulant matrices also have certain properties, such as commuting with each other and having orthogonal eigenvectors, that make them useful in various applications.

### Q: How are circulant matrices related to the eigenvectors of the permutation matrix?

Permutation matrices can be expressed as a combination of powers of the permutation matrix, and thus share the same eigenvectors. The eigenvector matrix for all circulant matrices of size N is the same as the eigenvector matrix for the permutation matrix of size N. The eigenvectors of circulant matrices are orthogonal and have specific patterns based on the eigenvalues.

## Summary & Key Takeaways

• Circulant matrices have a special form where each row or column is a cyclic shift of the previous row or column.

• These matrices are closely connected to the discrete Fourier transform, which is a key algorithm in engineering and mathematics.

• Circulant matrices are used in applications such as machine learning and image processing to reduce the dimensionality of data and simplify computations.