Operator Norm (Induced Norm) If we think of matrices as "maps" that transform one vector to another, what relation does the input and output norms have ||A||p ? max (||Ax||p / ||x||p) = max (||Ax||p / ||x||p) ||x||p > 0 ||x||p = 1 The above definition implies that y = Ax ? ||y||p ? ||A||p ||x||p HW: Prove ||AB||p ? ||A||p ||B||p HW: Prove the right hand side equality in the above definition HW: Prove operator norm satisfies the 4 conditions in the definition of
Added by Joseph A.
Close
Step 1
We know that for any vector x, |Ax|_p <= |A|_p |x|_p (this is the definition of the operator norm). Therefore, we have: |ABx|_p = |A(Bx)|_p <= |A|_p |Bx|_p <= |A|_p |B|_p |x|_p Taking the supremum over all x with |x|_p = 1, we get: |AB|_p = sup{|ABx|_p : |x|_p Show more…
Show all steps
Your feedback will help us improve your experience
Adi S and 96 other Calculus 3 educators are ready to help you.
Ask a new question
Labs
Want to see this concept in action?
Explore this concept interactively to see how it behaves as you change inputs.
Key Concepts
Recommended Videos
Problem Statement: For an m x n matrix A, prove using the definition provided above that the spectral norm is given by ||A|| = max ||Ax|| / ||x||. Conclude that for any n x 1 vector x, ||Ax|| <= ||A|| ||x||. Using the conclusion above, prove that for conformable matrices A and B (AB is well defined), ||AB|| <= ||A|| ||B||.
Vincenzo Z.
Norm of matrix-vector product. Suppose A is an m x n matrix and x is an n-vector. A famous inequality relates ||x||, ||A||, and ||Ax||: ||Ax|| ≤ ||A||||x||. The left-hand side is the (vector) norm of the matrix-vector product; the right-hand side is the (scalar) product of the matrix and vector norms. Show this inequality. Hints. Let a_i^T be the ith row of A. Use the Cauchy–Schwarz inequality to get (a_i^T x)^2 ≤ ||a_i||^2||x||^2. Then add the resulting m inequalities.
Krishna S.
Norm Preservation Lemma: Unitary matrices preserve the Euclidean norm ||Ux|| = ||x||, ||UA|| = ||A|| Proof: ||Ux||^2 = (Ux)*TUx = x*TU*TUx = x*T(U*TU)x = x*T(U^-1U)x = x*TIx = ||x||^2 => ||Ux|| = ||x|| Prove ||UA|| = ||A|| yourself! Note: It follows that the norm of a unitary matrix is 1 ||U|| = max ||Ux|| / ||x|| = 1
Sri K.
Recommended Textbooks
Calculus: Early Transcendentals
Thomas Calculus
Transcript
Watch the video solution with this free unlock.
EMAIL
PASSWORD