arXiv research

Study isotropy groups for complex orthogonal and skew-symmetric matrices.

Algorithm finds isotropy subgroups of orthogonal similarity on symmetric matrices.

We construct a decomposition of the identity operator on a Riemannian manifold $M$ as a sum of smooth orthogonal projections subordinate to an open cover of $M$. This extends a decomposition of the real line by smooth orthogonal projection due to Coifman, Meyer and Auscher, Weiss, Wickerhauser, and a similar decomposit…

Improves model predictability by mixing forecasts and orthogonalizing models.

A new hashing method improves accuracy by learning an orthogonal transform.

fMRI is a unique non-invasive approach for understanding the functional organization of the human brain, and task-based fMRI promotes identification of functionally relevant brain regions associated with a given task. Here, we use fMRI (using the Poffenberger Paradigm) data collected in mono- and dizygotic twin pairs t…

New convergence guarantees for learning with unknown nuisance parameters.

Orthogonal deep models defend against black-box attacks by ensuring internal representations are nearly orthogonal.

We construct an explicit topological model (similar to the topological Springer fibers appearing in work of Khovanov and Russell) for every two-row Springer fiber associated with the even orthogonal group and prove that the respective topological model is homeomorphic to its corresponding Springer fiber. This confirms …

Different neural networks trained on the same dataset often learn similar input-output mappings with very different weights. Is there some correspondence between these neural network solutions? For linear networks, it has been shown that different instances of the same network architecture encode the same representatio…

Special orthogonal representations from octonions have geometric properties linked to binary cubics.

Muon optimizer simplifies matrix optimization with spectral orthogonalization.

Random convolutional networks can be fooled with adversarial examples.

State-of-the-art algorithms for sparse subspace clustering perform spectral clustering on a similarity matrix typically obtained by representing each data point as a sparse combination of other points using either basis pursuit (BP) or orthogonal matching pursuit (OMP). BP-based methods are often prohibitive in practic…

We consider the problem of sampling from posterior distributions for Bayesian models where some parameters are restricted to be orthogonal matrices. Such matrices are sometimes used in neural networks models for reasons of regularization and stabilization of training procedures, and also can parameterize matrices of bo…

We classify six-dimensional Lie groups which admit a left-invariant half-flat SU(3)-structure and which split in a direct product of three-dimensional factors. Moreover, a complete list of those direct products is obtained which admit a left-invariant half-flat SU(3)-structure such that the three-dimensional factors ar…

Paper improves feature selection accuracy using transfer learning.

In this paper, we propose a scalable algorithm for spectral embedding. The latter is a standard tool for graph clustering. However, its computational bottleneck is the eigendecomposition of the graph Laplacian matrix, which prevents its application to large-scale graphs. Our contribution consists of reformulating spect…

Improved Gaussian process models for interpretable predictions.

Study asymptotics of extension and orthogonal Bergman kernels for high tensor powers of positive line bundles.

Anti-transfer learning prevents misleading representations for speech tasks.

Deterministic bounds for tensor singular values and vectors, differing from matrix cases.

Defines a similarity measure for classification distributions.

OGD proves robustness to Catastrophic Forgetting in Continual Learning.

The paper shows how gradient flow on over-parametrized tensor decomposition behaves like deflation.

ORFit trains models on streaming data with one pass, minimizing memory and computational costs.

Let $R$ be an infinite commutative ring with identity and $n\geq 2$ be an integer. We prove that for each integer $i=0,1,\cdots ,n-2,$ the $L^{2}$-Betti number $b_{i}^{(2)}(G)=0,$ $\$when $G=\mathrm{GL}_{n}(R)$ the general linear group, $\mathrm{SL}_{n}(R)$ the special linear group, $$ the group generated by…

New method improves reinforcement learning generalization.

Geometrically transforms word embeddings into a common space for better comparison.

Overparameterized models improve performance in sequential learning tasks.

In this paper we explore the "vector semantics" problem from the perspective of "almost orthogonal" property of high-dimensional random vectors. We show that this intriguing property can be used to "memorize" random vectors by simply adding them, and we provide an efficient probabilistic solution to the set membership …

Racah matrices and higher $j$-symbols are used in description of braiding properties of conformal blocks and in construction of knot polynomials. However, in complicated cases the logic is actually inverted: they are much better deduced from these applications than from the basic representation theory. Following the re…

Multi-head attention mechanism is capable of learning various representations from sequential data while paying attention to different subsequences, e.g., word-pieces or syllables in a spoken word. From the subsequences, it retrieves richer information than a single-head attention which only summarizes the whole sequen…

The performance of Orthogonal Matching Pursuit (OMP) for variable selection is analyzed for random designs. When contrasted with the deterministic case, since the performance is here measured after averaging over the distribution of the design matrix, one can have far less stringent sparsity constraints on the coeffici…

It is well known that the curvature tensor of a pseudo-Riemannian manifold can be decomposed with respect to the pseudo-orthogonal group into the sum of the Weyl conformal curvature tensor, the traceless part of the Ricci tensor and of the scalar curvature. A similar decomposition with respect to the pseudo-unitary gro…

String structures have played an important role in algebraic topology, via elliptic genera and elliptic cohomology, in differential geometry, via the study of higher geometric structures, and in physics, via partition functions. We extend the description of String structures from connected covers of the definite-signat…

In this paper, we present new results on using orthogonal matching pursuit (OMP), to solve the sparse approximation problem over redundant dictionaries for complex cases (i.e., complex measurement vector, complex dictionary and complex additive white Gaussian noise (CAWGN)). A sufficient condition that OMP can recover …

Lipschitz constraints under L2 norm on deep neural networks are useful for provable adversarial robustness bounds, stable training, and Wasserstein distance estimation. While heuristic approaches such as the gradient penalty have seen much practical success, it is challenging to achieve similar practical performance wh…

New exponential map for Lie groups connects to sub-Riemannian geometry.

Distance metric learning (DML), which learns a distance metric from labeled "similar" and "dissimilar" data pairs, is widely utilized. Recently, several works investigate orthogonality-promoting regularization (OPR), which encourages the projection vectors in DML to be close to being orthogonal, to achieve three effect…

New methods calibrate causal estimates using standard predictive models.

The study extends Jacobi-orthogonality to indefinite scalar product spaces.

New characterization of Osserman tensors using Jacobi-orthogonality.

New findings on Kähler manifolds restrict orthogonal coordinates existence.

Constructs orthogonal coordinates in curved spaces.

OPT framework improves neural network generalization by learning an orthogonal transformation.

The paper studies surfaces in a bounded domain with orthogonal boundaries and proves curvature estimates.

Orthogonal random features approximate a Bessel kernel, offering sharper bounds than random Fourier features.