Questions tagged [tensor-decomposition]
The tensor-decomposition tag has no usage guidance.
95
questions
1
vote
0answers
17 views
What is the maximum of $\operatorname{Tr}(EUAU^\dagger)$ when $U$ is unitary?
Given matrices $E$ and $A$, is there a good way to obtain the maximizing unitary $U$ for the objective function $\operatorname{Tr}(EUAU^*)$? (I know that the maximizer of $\operatorname{Tr}(EU)$ is ...
0
votes
0answers
38 views
Special case of CP decomposition
I am reading this article by Kolda. On page 21, it says:
...CP [decomposition] can be viewed as a special case of Tucker [decomposition] where the core tensor is superdiagonal and $P = Q = R$...
...
3
votes
0answers
61 views
An svd-like tensor decomposition splitting a tensor into two lower-dimensional tensors and a singular vector
I have also posted this question on mathoverflow, but it seems there are a lot of questions related to SVDs here and a tag "tensor-decomposition", so I will give it a shot.
I am looking for ...
2
votes
1answer
41 views
Determine wether a skew-symmetric tensor is decomposable or not.
For $x_1,x_2,x_3,x_4$ elements of a vector space $V$, I would like to find out wheter the skew-symmetric tensor $11x_1 \wedge x_2 + 10x_1 \wedge x_3 + 17x_1 \wedge x_4 − 13x_2 \wedge x_3 − 10x_2 \...
0
votes
0answers
17 views
How does the outer product of matrics work during approximation of a tensor?
I read a paper about Tensor-SVD, named Factorization strategies for third-order tensors.
The section 4.1 provides an approximation strategies based on products of tensors, which is shown as follows:
...
0
votes
0answers
18 views
rank of a matrix formed from a tensor product
I know that rank of a matrix expressed in term of product of 2 matrices, satisfies the following equation rank(AB) ≤ min(rank(A), rank(B)).
When we write a matrix as a product of tensors(of tensor ...
0
votes
0answers
9 views
Single-factor CP decomposition optimization equivalency with block coordinate descent using the tensor power method
I am trying to understand the objective functions for CP decompositions using the tensor power method as discussed in this paper. In this approach, a series of rank-1 approximations are made using the ...
0
votes
0answers
11 views
Covariance of Wishart Distribution
Let us consider a random matrix $$W \sim Wishart(n, \Sigma_p),$$ i.e. $W \in \mathbb{R}^{p \times p}$ and $W = Z^T Z$ for $Z \in \mathbb{R}^{n \times p}$ and each row of $Z$ drawn iid from a $N(0, \...
0
votes
0answers
26 views
The optimal solution of tensor Tucker decomposition problem
I have a question about tensor Tucker deocomposition. Recall that in Higher-Order Singular Value Decomposition (HOSVD), each factor matrices $U_i$ is obtained by taking the top $R_i$ left singular ...
0
votes
1answer
28 views
How to build 2nd rank tensors from set of vectors?
Suppose we have 3 vectors: $A_i$, $B_i$, $C_i$, where $i=1,2,3$. How to build all possible dependent 2nd rank tensors based on these vectors?
I believe, that at first we need to obtain all possible ...
0
votes
0answers
11 views
Choosing an Irreducible Tensor Operator Basis where the Singular Values of Each Basis Element are the Same
Let $\mathcal{B(H)}$ be the space of all bounded linear operators on the Hilbert space $\mathcal{H}$. Let $g \rightarrow \mathcal{U}_g$, where $\mathcal{U}_g (\cdot):= U(g) (\cdot) U(g)^\dagger $, be ...
0
votes
0answers
17 views
What's about a tensor/matrix decomposition that makes it better?
The following question is for tensors, but could also be applied to matrices.
I have seen that an N-th order tensor $X \in \mathbb{R}^{I\times...\times I}$ has an original storage complexity of $O(I^N)...
5
votes
1answer
118 views
Schouten theorem and doubts in the calculation of the tensor derivative
I am working on the following theorem
Theorem (J. A. Schouten)
For $n\geq 4$ the metric $g$ is conformally flat if and only if $W=0$. For $n=3$, the metric $g$ is conformally flat if and only if the ...
0
votes
1answer
47 views
References to Q-curvature
I am working on Q-curvature:
$$
Q^n_g= \frac{1}{2(n-1)} \Delta_g
R_g+\frac{n^3-4n^2+16n-16}{8(n-1)^2(n-2)^2}\:R_g-\frac{2}{(n-2)^2}\:| Ric |^2_g
$$
See, for example, the expression (3) in https://...
0
votes
0answers
28 views
CP-ALS algorithm with Kroonenberg and De Leeuw initialization
I am trying to implement the CANDECOMP/PARAFAC - Alternating Least Square algorithm, with a small improvement in the initial guess. The simple initial guess for A is random in the method==0 if block. ...
1
vote
1answer
58 views
SVD in scipy and numpy for tensors
Can someone explain to me the difference between SVD of numpy and scipy for Multidimensional arrays (Tensors)?
...
0
votes
0answers
16 views
Decomposition of a matrix as a tensor product
Suppose that a squared-matrix $A$ can be decomposed as a tensor product of lower rank squared matrices, i.e.
$$
A = M_1\otimes M_2.
$$
For simplicity, suppose $A$ is a $N\times N$ matrix where $N/2$ ...
0
votes
0answers
32 views
Block metric tensor and Fermi coordinates.
Consider a manifold $M^2\subset\mathbb{R}^4$ locally parametrized by $Y(\xi^1,\xi^2)$ and let $X$ be a Fermi coordinates parametrization of a neighbourhood of $M$ (which we call $\bar{M}$), that is
$$
...
1
vote
2answers
49 views
Why in tensors is the identity $\int{A_{\nu}\partial_{\mu}}\partial^{\nu}A^{\mu} = \int A_{\mu}\partial^{\mu}\partial^{\nu}A_{\nu}$ valid
I am trying to understand the and there is a step in the Derivation of something and there is a step which I cannot understand.
Its basically a lack of understanding of tensors so any help would be ...
1
vote
0answers
17 views
Condition for endomorphism to be an element of tensor product of Lie representations
I'm coming from a physics background, so apologies if this question is technically ill-phrased. Consider an operator $H\in \text{End}(\mathcal{H}_{A}\otimes \mathcal{H}_{B})$, where $\mathcal{H}_{X}$ ...
0
votes
2answers
73 views
Proving $R/I \otimes_R M \cong_R M/IM$ using UMP and a biadditive map not with exact sequences.
I want to Understand a paragraph of the proof of $R/I \otimes_R M \cong_R M/IM.$ in example$(8)$ on pg. 370 of Dummit and Foote (third edition)
Here is the example from Dummit & Foote:
My ...
0
votes
0answers
18 views
What are the sufficient and necessary conditions for a unitary matrix to be written in a tensor product of 2x2 matrices?
I understand an arbitrary unitary matrix cannot be always written in a tensor product of 2x2 matrices. But, if a unitary matrix meets some conditions, can it be done? If so, what are those conditions?
6
votes
2answers
183 views
Is every harmonic polynomial a linear combination of these?
In $N$-dimensional space, we can show by direct calculation that the polynomial
$$
r^{2K+N-2}\nabla_a\nabla_b\nabla_c\cdots \frac{1}{r^{N-2}}
\hspace{1cm}
\text{(with $K$ derivatives)}
$$
is ...
3
votes
1answer
53 views
How to show this tensor is not decomposable
Let $V$ a vector space over $\mathbb{F}$ with basis $B=\{e_1,\dots, e_n \}$ and $B'=\{e^1,\dots,e^n \}$ adjoint basis in $V^*$, Consider the following tensor:
$$t^{ij}=(i+j) $$
I need to see that is ...
0
votes
0answers
34 views
How can I express a tensor in terms of a basis different to canonical?
Considerthe canonical basis $(e_1,e_2,e_3)$ and its dual basis $e^1 , e^2 , e^3$. Let $A = (a_{ij})$ a matrix in $M_{3 \times 3}(\mathbb{R})$ given by:
$$A = \begin{bmatrix}
1 & -3 & 1\\
-3 &...
4
votes
2answers
179 views
A tensor over $\mathbb R$ is sum of two decomposable tensors
Let $W$ be a vector space over $\mathbb R$ with basis $\{a,b\}$. Consider the tensor
$$a \otimes a \otimes a - b \otimes b \otimes a + a \otimes b \otimes b + b \otimes a \otimes b.$$
a) Show that ...
0
votes
0answers
17 views
Component of a irreducible tensor product
Given the operators $\boldsymbol{\alpha}$ and $\boldsymbol{C^{(L)}}$ such that
$$
\boldsymbol{\alpha}=\left(\begin{array}{cc}
0 & \boldsymbol{\sigma}_{p} \\
\boldsymbol{\sigma}_{p} & 0
\end{...
1
vote
1answer
34 views
Tensor Notation ambiguity
Sorry for this silly question, but this notation keeps popping up in my face in too many different places without any clarification, so I will be so grateful if someone can help and answer my question:...
1
vote
0answers
61 views
Frobenius norm optimization problem
How to solve the following optimization problem which arises in the field of Higher Order SVD?
Let $A \in \mathbb{R}^{m\times n}$, $W \in \mathbb{R}^{m\times l}$.
find $\arg \max _{A}\|A^T W\|_F$ s.t. ...
0
votes
0answers
25 views
Dimensionalities of inner Product of tensors
As the inner product of two vectors i.e. tensors of order $1$ results into a reduced-order tensor i.e. scalar (a tensor of order $0$).
as the outer product of two vectors results into a matrix. while ...
2
votes
1answer
60 views
How to algebraically express the condition for a tensor to be of rank $1$?
Let $W$ be a finite-dimensional complex vector space and let $V = W\otimes \cdots \otimes W$ ($m$ times). A non-zero element $v \in V$ is said to be of rank $1$ if $v$ can be written as
$$v = w_1 \...
0
votes
1answer
58 views
How many projections $\pi: \mathbb{F}_3^2 \to \mathbb{F}_3^2 $ there? [closed]
How many projections $\pi: \mathbb{F}_3^2 \to \mathbb{F}_3^2 $ there?
Justify your answer!
If someone can give me a tipps on this question.D
2
votes
1answer
319 views
Irreducible representation of $SO(3)$ by rank-3 tensors and higher?
I read that any rank-2 tensor can be decomposed into the sum of a traceless symmetric tensor, an anti-symmetric tensor and a unit tensor, all closed under $SO(3)$. The three form an irreducible ...
1
vote
1answer
47 views
Given $\sum_i\sum_j \sigma_{ij} a_i b_j^T$ and orthogonal $\{a_i\}$, find orthogonal $\{b_j\}$
Suppose I have a matrix $T \in \mathbb{R}^{n \times n}$ of the form:
$$T = \sum_{i = 1}^{n_A} \sum_{j = 1}^{n_B} \sigma_{ij} a_i b_j^T$$
where $1 \leq n_A, n_B \leq n$, $\sigma_{ij} \in \mathbb{R}\...
0
votes
1answer
28 views
Integrating unitary vector in calculating stress resultant
The problem defines:
velocity in the local base $\{e_r , e_\theta , e_z\}$, the gradient
and acceleration in the local curvilignear base $e_{i} \otimes e_{j}$
avec $i, j \in\{r, \theta, z\}$
...
0
votes
1answer
33 views
Find orthonormal $\{b_i\}$ such that $\mathcal{A} = \sum_i \sum_j \lambda_i \mu_j b_i b_j^T$
Suppose $\{b_i\}_{i = 1}^d \subset \mathbb{R}^n$ is an orthonormal set of vectors, $d \leq n$, and assume that a matrix $\mathcal{A} \in \mathbb{R}^{n \times n}$ has the following form:
$$\mathcal{A} ...
0
votes
0answers
46 views
Gradient Descent for Tensor Decomposition - Find low rank dimensions
I encountered this paper on travel time estimation using Tensor Decomposition and at Page 4, Figure 5 there is an algorithm to decompose a tensor using Gradient Descent.
The first line of this ...
0
votes
1answer
38 views
linear independence of symmetric tensors
I am reading a paper that incidentally uses a bit of theory of symmetric tensor spaces. I came across the following claim:
If we're given linearly independent vectors
$x_1, \ldots, x_n \in \mathbb{...
1
vote
0answers
16 views
Tucker and TT decompositions in 3D case
I’m wondering if Tucker factorization and Tensor train decomposition are equivalent in case of 3-rd order tensor and how to show it?
2
votes
1answer
135 views
Decomposition of tensor product of defining representation with itself for $G=\mathrm{SO}(5)$
I want to decompose the tensor product $V\otimes V$, where $V=\Bbb C^5$ denotes the defining representation of $\mathrm{SO}(5)$, into irreducible representations using the following formula for the ...
2
votes
1answer
112 views
Finding the irreducible components of a rank 3 tensor
In 3 dimensions, a rank-2 tensor can be identified via a scalar, a vector and a (symmetric, traceless) tensor component by contracting it with $\delta_{ij}$ or $\epsilon_{ijk}$:
$$ T_{ij} = \...
3
votes
1answer
117 views
Proof of 3-fold Tensor Uniqueness
In the notes linked below Professor Roughgarden states Theorem 3.1 without proof.
http://timroughgarden.org/s17/l/l10.pdf
Any ideas how one would go about proving this statement or references to ...
1
vote
1answer
185 views
Tensor product raised to power $N$.
Can the following quantity be reduced to further
$[A \otimes B + (\mathbb{1} - A) \otimes D]^N$, for positive interger $N$? Here, $\otimes$ denotes the Kronecker (tensor) product and the matrices $A$ ...
1
vote
0answers
111 views
What are the matricization and vectorization of tensor products?
I'm trying to understand the concepts of matricization (matrix unfolding) and vectorization of tensor products. In the past, I've only dealt with tensor products of infinite-dimensional Banach and ...
1
vote
2answers
77 views
If a tensor's multilinear rank is $(R,R,R)$, then is its canonical/CP rank also $R$?
Given an order $2$ tensor (i.e. a matrix), one always has that row rank is equal to the column rank, so that its multilinear rank or $n$-ranks is always equal to $(R,R)$ for some $R$. Moreover that $R$...
0
votes
1answer
61 views
non totally decomposability of vectors in Grassmannian
In "Algebraic geometry: a first course", by Harris, Grassmannian is described, under the Plucker embedding, as the locus of totally decomposable vectors in the projectivization of the exterior power $\...
2
votes
0answers
28 views
What are the irreducible representations of $\mathbf{10}\otimes \mathbf{3}$ in $SU(3)$
Since, a rank-3 tensor has 10 components and a rank-1 tensor has 3 components in $SU(3)$, I know that we are searching for the different irreducible representations of the tensor $v_{ijk}w_{l}$.
The ...
4
votes
0answers
56 views
Exterior power of symmetric product of GL(2,R) tensor representations
A paper I am reading uses an identification of $GL(2,\mathbb{R})$ representations:
\begin{align}
\Lambda^2(\text{Sym}^3\mathbb{R}^2) = (\text{Sym}^4\mathbb{R}^2 \otimes \Lambda^2\mathbb{R}^2 ) \...
0
votes
1answer
297 views
Maximal Ideals in Tensor product of Algebras
Let A, B be algebras over algebraically closed field $\mathbb F$, and m be a maximal ideal in $A\otimes_\mathbb{F} B$ such that $A\otimes_\mathbb{F} B/m \cong \mathbb F$. Show that there are maximal ...
1
vote
1answer
88 views
Readings on Tensors, Tensor Algebra and Tensor Decomposition
I have just started my Masters Degree in Mathematics and I will be focussing on Tensors (Viewed as Multidimensional Arrays) and Tensor decompositions.
My professor is by no means an expert on Tensors ...
