New upper bounds for the spectral variation of a general matrix

Abstract: Let

A i n m a t h b b C^{n i m e s n}

be a normal matrix with spectrum

{l a m b d a_{i}}_{i = 1}^{n}

, and let

w i d e t i l d e A = A + E i n m a t h b b C^{n i m e s n}

be a perturbed matrix with spectrum

{w i d e t i l d e {l a m b d a}_{i}}_{i = 1}^{n}

. If

w i d e t i l d e A

is still normal, the celebrated Hoffman--Wielandt theorem states that there exists a permutation

p i

of

1, l d o t s, n

such that

, where

| c d o t |_{F}

denotes the Frobenius norm of a matrix. This theorem reveals the strong stability of the spectrum of a normal matrix. However, if

A

or

w i d e t i l d e A

is non-normal, the Hoffman--Wielandt theorem does not hold in general. In this paper, we present new upper bounds for

, provided that both

A

and

w i d e t i l d e A

are general matrices. Some of our estimates improve or generalize the existing ones.

Recommendations

Cites work

Cited in

(3)

This page was built for publication: New upper bounds for the spectral variation of a general matrix