Source author record

Ali Zamani

Explore this author’s publications, collaborators and research topics. This catalog record has not been claimed by its author.

ResearcherUnclaimed source record

Catalog footprint

Research overview

8works
2topics
4close collaborators

Published work

8 publications

preprint2022arXiv

A geometric approach to inequalities for the Hilbert--Schmidt norm

We define angle ΘX,YΘ_{X,Y} between non-zero Hilbert--Schmidt operators XX and YY by cos⁡ΘX,Y=ReTr(Y∗X)∥X∥2∥Y∥2\cosΘ_{_{X,Y}} = \frac{{\rm Re}{\rm Tr}(Y^*X)}{{\|X\|}_{_2}{\|Y\|}_{_2}}, and give some of its essentially properties. It is shown, among other things, that \begin{align*} \big|\cosΘ_{_{X,Y}}\big|\leq \min\left\{\sqrt{\cosΘ_{_{|X^*|,|Y^*|}}}, \sqrt{\cosΘ_{_{|X|,|Y|}}}\right\}. \end{align*} It enables us to provide alternative proof of some well-known inequalities for the Hilbert--Schmidt norm. In particular, we apply this inequality to prove Lee's conjecture [Linear Algebra Appl. 433 (2010), no.~3, 580--584] as follows \begin{align*} {\big\|X + Y\big\|}_{_2} \leq \sqrt{\frac{\sqrt{2} + 1}{2}}\,{\big\|\,|X| + |Y|\,\big\|}_{_2}. \end{align*} A numerical example is presented to show the constant 2+12\sqrt{\frac{\sqrt{2} + 1}{2}} is smallest possible. Other related inequalities for the Hilbert--Schmidt norm are also considered.

preprint2022arXiv

C*-module operators which satisfy in the generalized Cauchy--Schwarz type inequality

Let L(H)\mathcal{L}(\mathscr{H}) denote the C∗C^*-algebra of adjointable operators on a Hilbert C∗C^*-module H\mathscr{H}. We introduce the generalized Cauchy-Schwarz inequality for operators in L(H)\mathcal{L}(\mathscr{H}) and investigate various properties of operators which satisfy the generalized Cauchy--Schwarz inequality. In particular, we prove that if an operator A∈L(H)A\in\mathcal{L}(\mathscr{H}) satisfies the generalized Cauchy-Schwarz inequality such that AA has the polar decomposition, then AA is paranormal. In addition, we show that if for AA the equality holds in the generalized Cauchy-Schwarz inequality, then AA is cohyponormal. Among other things, when AA has the polar decomposition, we prove that AA is semi-hyponormal if and only if ∥⟨Ax,y⟩∥≤∥∣A∣1/2x∥∥∣A∣1/2y∥\big\|\langle Ax, y\rangle\big\| \leq \big\|{|A|}^{1/2}x\big\|\big\|{|A|}^{1/2}y\big\| for all x,y∈Hx, y \in\mathscr{H}.

preprint2020arXiv

A-numerical radius and product of semi-Hilbertian operators

Let AA be a positive bounded operator on a Hilbert space (H,⟨⋅,⋅⟩)\big(\mathcal{H}, \langle \cdot, \cdot\rangle \big). The semi-inner product ⟨x,y⟩A:=⟨Ax,y⟩{\langle x, y\rangle}_A := \langle Ax, y\rangle, x,y∈H,x, y\in\mathcal{H}, induces a seminorm ∥⋅∥A{\|\cdot\|}_A on H\mathcal{H}. Let wA(T)w_A(T) denote the AA-numerical radius of an operator TT in the semi-Hilbertian space (H,∥⋅∥A)\big(\mathcal{H}, {\|\cdot\|}_A\big). In this paper, for any semi-Hilbertian operators TT and SS, we show that wA(TR)=wA(SR)w_A(TR) = w_A(SR) for all (AA-rank one) semi-Hilbertian operator RR if and only if A1/2T=λA1/2SA^{1/2}T = λA^{1/2}S for some complex unit λλ. From this result we derive a number of consequences.

preprint2020arXiv

Numerical radius orthogonality in C∗C^*-algebras

In this paper we characterize the Birkhoff--James orthogonality with respect to the numerical radius norm v(⋅)v(\cdot) in C∗C^*-algebras. More precisely, for two elements a,ba, b in a C∗C^*-algebra A\mathfrak{A}, we show that a⊥Bvba\perp_{B}^{v} b if and only if for each θ∈[0,2π)θ\in [0, 2π), there exists a state φθφ_{_θ} on A\mathfrak{A} such that ∣φθ(a)∣=v(a)|φ_{_θ}(a)| = v(a) and $\mbox{Re}\big(e^{iθ}\overline{φ_{_θ}(a)}φ_{_θ}(b)\big)\geq 0$. Moreover, we compute the numerical radius derivatives in A\mathfrak{A}. In addition, we characterize when the numerical radius norm of the sum of two (or three) elements in A\mathfrak{A} equals the sum of their numerical radius norms.

preprint2020arXiv

On an extension of operator transforms

We introduce the λλ-mean transform Mλ(T)M_λ(T) of a Hilbert space operator TT as an extension of some operator transforms based on the Duggal transform TDT^D by Mλ(T):=λT+(1−λ)TDM_λ(T) := λT + (1-λ)T^D, and present some of its essentially properties. Among other things, we obtain estimates for the operator norm and numerical radius of the λλ-mean transform Mλ(T)M_λ(T) in terms of the original operator TT.

preprint2020arXiv

Some upper bounds for the A\mathbb{A}-numerical radius of 2×22\times 2 block matrices

Let A=(A00A)\mathbb{A}=\left( \begin{array}{cc} A & 0 \\ 0 & A \\ \end{array} \right) be the 2×22\times2 diagonal operator matrix determined by a positive bounded operator AA. For semi-Hilbertian operators XX and YY, we first show that \begin{align*} w^2_{\mathbb{A}}\left(\begin{bmatrix} 0 & X \\ Y & 0 \end{bmatrix}\right) &\leq \frac{1}{4}\max\Big\{{\big\|XX^{\sharp_A} + Y^{\sharp_A}Y\big\|}_{A}, {\big\|X^{\sharp_A}X + YY^{\sharp_A}\big\|}_{A}\Big\} + \frac{1}{2}\max\big\{w_{A}(XY), w_{A}(YX)\big\}, \end{align*} where wA(⋅)w_{\mathbb{A}}(\cdot), ∥⋅∥A{\|\cdot\|}_{A} and wA(⋅)w_{A}(\cdot) are the A\mathbb{A}-numerical radius, AA-operator seminorm and AA-numerical radius, respectively. We then apply the above inequality to find some upper bounds for the A\mathbb{A}-numerical radius of certain 2×22\times 2 operator matrices. In particular, we obtain some refinements of earlier AA-numerical radius inequalities for semi-Hilbertian operators. An upper bound for the A\mathbb{A}-numerical radius of 2×22\times 2 block matrices of semi-Hilbertian space operators is also given.

preprint2016arXiv

Mappings preserving approximate orthogonality in Hilbert C∗C^*-modules

We introduce a notion of approximate orthogonality preserving mappings between Hilbert C∗C^*-modules. We define the concept of (δ,ε)(δ, \varepsilon)-orthogonality preserving mapping and give some sufficient conditions for a linear mapping to be (δ,ε)(δ, \varepsilon)-orthogonality preserving. In particular, if E\mathscr{E} is a full Hilbert A\mathscr{A}-module with K(H)⊆A⊆B(H)\mathbb{K}(\mathscr{H})\subseteq \mathscr{A} \subseteq \mathbb{B}(\mathscr{H}) and T,S:E⟶ET, S:\mathscr{E}\longrightarrow \mathscr{E} are two linear mappings satisfying ∣⟨Sx,Sy⟩∣=∥S∥2 ∣⟨x,y⟩∣\big|\langle Sx, Sy\rangle\big| = \|S\|^2\,|\langle x, y\rangle| for all x,y∈Ex, y\in \mathscr{E} and ∥T−S∥≤θ∥S∥\|T - S\| \leq θ\|S\|, then we show that TT is a (δ,ε)(δ, \varepsilon)-orthogonality preserving mapping. We also prove whenever K(H)⊆A⊆B(H)\mathbb{K}(\mathscr{H})\subseteq \mathscr{A} \subseteq \mathbb{B}(\mathscr{H}) and T:E⟶FT: \mathscr{E} \longrightarrow \mathscr{F} is a nonzero A\mathscr{A}-linear (δ,ε)(δ, \varepsilon)-orthogonality preserving mapping between A\mathscr{A}-modules, then ∥⟨Tx,Ty⟩−∥T∥2⟨x,y⟩∥≤4(ε−δ)(1−δ)(1+ε)∥Tx∥ ∥Ty∥(x,y∈E).\big\|\langle Tx, Ty\rangle - \|T\|^2\langle x, y\rangle\big\|\leq \frac{4(\varepsilon - δ)}{(1 - δ)(1 + \varepsilon)} \|Tx\|\,\|Ty\|\qquad (x, y\in \mathscr{E}). As a result, we present some characterizations of the orthogonality preserving mappings.

preprint2015arXiv

Approximately bisectrix-orthogonality preserving mappings

Regarding the geometry of a real normed space X{\mathcal X}, we mainly introduce a notion of approximate bisectrix-orthogonality on vectors x,y∈Xx, y \in {\mathcal X} as follows: $${x\np{\varepsilon}}_W y \mbox{if and only if} \sqrt{2}\frac{1-\varepsilon}{1+\varepsilon}\|x\|\,\|y\|\leq \Big\|\,\|y\|x+\|x\|y\,\Big\|\leq\sqrt{2}\frac{1+\varepsilon}{1-\varepsilon}\|x\|\,\|y\|.$$ We study class of linear mappings preserving the approximately bisectrix-orthogonality ${\np{\varepsilon}}_W$. In particular, we show that if T:X→YT: {\mathcal X}\to {\mathcal Y} is an approximate linear similarity, then $${x\npδ}_W y\Longrightarrow {Tx \npθ}_W Ty \qquad (x, y\in {\mathcal X})$$ for any δ∈[0,1)δ\in[0, 1) and certain θ≥0θ\geq 0.

Research graph

See the researcher in context

Open full explorer

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.