Scribd
Upload
136 views9 pages

Matrix Inversion Lemma Explained

This document derives two matrix identities and uses them to obtain expressions for the inverse of (A + BCD). The expressions, known as the Matrix Inversion Lemma or Sherman-Morrison-Woodbury Identity, allow calculating the inverse by using the inverse of A and the terms B, C, and D. Two special cases are noted as useful for relating forms of the Kalman filter that estimate state and covariance over time.

Uploaded by

Hunmin Shin
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
136 views9 pages

Matrix Inversion Lemma Explained

This document derives two matrix identities and uses them to obtain expressions for the inverse of (A + BCD). The expressions, known as the Matrix Inversion Lemma or Sherman-Morrison-Woodbury Identity, allow calculating the inverse by using the inverse of A and the terms B, C, and D. Two special cases are noted as useful for relating forms of the Kalman filter that estimate state and covariance over time.

Uploaded by

Hunmin Shin
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Matrix Inversion Identities

Ged Ridgway Centre for Medical Image Computing University College London

November, 2006

Summary

Two simple matrix identities are derived, these are then used to get expressions for the inverse of (A + BCD). The expressions are variously known as the Matrix Inversion Lemma or Sherman-Morrison-Woodbury Identity. The derivation in these slides is taken from Henderson and Searle [1]. An alternative derivation, leading to a generalised expression, can be found in Tylavsky and Sohie [2]. Two special case results are mentioned, as they are useful in relating the Kalman-gain form and Information form of the Kalman Filter.

Identity 1

(I + P)1 = (I + P)1 (I + P P) = I (I + P)1 P (1)

Identity 2

P + PQP = P(I + QP) = (I + PQ)P (I + PQ)1 P = P(I + QP)1 (2)

Matrix Inversion Lemma - step 1

For invertible A, but general (possibly rectangular) B,C, and D: (A + BCD)1 = A I + A1 BCD = I + A1 BCD
1 1

A1
1

= I I + A1 BCD

A1 BCD A1

Using (1)

= A1 (I + A1 BCD)1 A1 BCDA1

Matrix Inversion Lemma - step 2


Repeatedly using (2) in sequence now produces: (A + BCD)1 = A1 (I + A1 BCD)1 A1 BCDA1 =A =A =A
1 1

(3) (4) (5) (6) (7) (8)

A A

1 1

(I + BCDA

1 1

BCDA

1 1

B(I + CDA

B)

CDA

= A1 A1 BC(I + DA1 BC)1 DA1


1

BCD(I + A

BCD)

1 1

= A1 A1 BCDA1 (I + BCDA1 )1 (note that the order ABCD is maintained, ignoring the other parts of the expressions)

Matrix Inversion Lemma - special case

If C is also invertible, from (5): (A + BCD)1 = A1 A1 B(I + CDA1 B)1 CDA1 = A1 A1 B(C 1 + DA1 B)1 DA1 (9)

which is a commonly used variant (for example applicable to the Kalman Filter covariance, in the correction step of the lter).

Another related special case

A very similar use of (2) gives: (A + BCD)1 BC = A1 (I + BCDA1 )1 BC = A1 B(I + CDA1 B)1 C and for invertible C: = A1 B(C 1 + DA1 B)1 which is useful in converting between Kalman-gain and Information forms of the Kalman Filter state-estimate correction step. (10) (11)

Bibliography

H. V. Henderson and S. R. Searle. On Deriving the Inverse of a Sum of Matrices. SIAM Review, 23(1):5360, January 1981. D.J. Tylavsky and G.R.L. Sohie. Generalization of the matrix inversion lemma. Proceedings of the IEEE, 74(7):10501052, July 1986.

You might also like