arXiv research

A new probabilistic BTD method for tensor data.

Combining neural networks and multiscale decomposition for financial market analysis.

Structured linear substitutions improve both efficiency and accuracy in neural networks.

Adaptive tensor modeling preserves continuity in multidimensional data.

The purpose of this paper is to describe certain natural 4-vector fields on quaternionic flag manifolds, which geometrically determine the Bruhat cell decomposition. This structure naturally descends from the symplectic group, where it is related to the dressing action given by the Iwasawa decomposition of the general …

SKOLR uses linear RNNs to approximate Koopman operators for time-series forecasting.

The theory of geometric structures on a surface with nonempty boundary can be developed by using a decomposition of such a surface into hexagons, in the same way as the theory of geometric structures on a surface without boundary is developed using the decomposition of such a surface into pairs of pants. The basic elem…

The space of probability densities is an infinite-dimensional Riemannian manifold, with Riemannian metrics in two flavors: Wasserstein and Fisher--Rao. The former is pivotal in optimal mass transport (OMT), whereas the latter occurs in information geometry---the differential geometric approach to statistics. The Rieman…

LLT transforms time series features based on linear laws.

Paper introduces Modular Jets for diagnosing model decompositions in pipelines.

In this letter, first we give a decomposition for any Lie-Poisson structure $π_g$ associated to the modular vector. In particular, $π_g$ splits into two compatible Lie-Poisson structures if $dim{g} \leq 3$. As an application, we classified quadratic deformations of Lie-Poisson structures on $\mathbb R^3$ up to linear d…

Galerkin method outperforms graph-based methods in spectral decompositions.

NA$_0$CT$^2$ improves tensor regression predictions with $\ell_0$ regularization.

Sparse incidence tensors can represent a variety of structured data. For example, we may represent attributed graphs using their node-node, node-edge, or edge-edge incidence matrices. In higher dimensions, incidence tensors can represent simplicial complexes and polytopes. In this paper, we formalize incidence tensors,…

Review of algorithms for linear system approximations.

Enhances forecasting of complex systems using FKMD.

Neural Decomposition breaks down VAE latent structure for better interpretability.

Extends matrix factorization for deviance-based losses with GLM theory.

Proposes MVGPR for spatiotemporal data modal analysis.

Two-dimensional embeddings remain the dominant approach to visualize high dimensional data. The choice of embeddings ranges from highly non-linear ones, which can capture complex relationships but are difficult to interpret quantitatively, to axis-aligned projections, which are easy to interpret but are limited to biva…

Study on identifying AMP chain graph models under known and unknown component decompositions.

In this article we describe cell decompositions of the moduli space of Riemann surfaces and their relationship to a Hurwitz problem. The cells possess natural linear structures and with respect to this they can be described as rational convex polytopes which come equipped with natural integer points and a volume form. …

Estimates MLDS using tensor decomposition, improving upon existing methods.

Quantization on even-dimensional compact manifolds using cell decomposition.

SVD improves neural network optimization.

We investigate aspects of semimartingale decompositions, approximation and the martingale representation for multidimensional correlated Markov processes. A new interpretation of the dependence among processes is given using the martingale approach. We show that it is possible to represent, in both continuous and discr…

Study the structure of equidistant decompositions in manifolds.

New algorithm improves dynamic mode decomposition for high-dimensional data.

Normal forms and moduli stacks for flat connections on complex manifolds.

Proposes a faster Isomap algorithm by reducing eigenvalue decomposition complexity.

Tensor decomposition recovers Gaussian mixtures from moments.

New tensor model reduces GLM estimation error and sample complexity.

New approach learns mixtures of linear dynamical systems without separation conditions.

Tensor decomposition is an important technique for capturing the high-order interactions among multiway data. Multi-linear tensor composition methods, such as the Tucker decomposition and the CANDECOMP/PARAFAC (CP), assume that the complex interactions among objects are multi-linear, and are thus insufficient to repres…

Study geometric flows of G2-structures, determining curvature and torsion invariants.

New algorithm for online tensor factorization with provable guarantees.

The paper uses tensor decompositions to improve neural network models for tree data.

Common positive stabilisation found for isotopic contact structures.

New method identifies latent variables with causal dependencies from observed data.

In this text we give a decomposition result on polynomial poly-vector fields generalizing a result on the decomposition of homogeneous Poisson structures. We discuss consequences of this decomposition result in particular for low dimensions and low degrees. We provide the tools to calculate simple cubic Poisson structu…

We present a simple, general technique for reducing the sample complexity of matrix and tensor decomposition algorithms applied to distributions. We use the technique to give a polynomial-time algorithm for standard ICA with sample complexity nearly linear in the dimension, thereby improving substantially on previous b…

Motivated by the algorithmic study of 3-dimensional manifolds, we explore the structural relationship between the JSJ decomposition of a given 3-manifold and its triangulations. Building on work of Bachman, Derby-Talbot and Sedgwick, we show that a "sufficiently complicated" JSJ decomposition of a 3-manifold enforces a…

We demonstrate the application of an algorithmic trading strategy based upon the recently developed dynamic mode decomposition (DMD) on portfolios of financial data. The method is capable of characterizing complex dynamical systems, in this case financial market dynamics, in an equation-free manner by decomposing the s…

ENTED efficiently decomposes binary and count tensors using nonparametric Gaussian processes.

We determine parts of the contact homology of certain contact 3-manifolds in the framework of open book decompositions, due to Giroux. We study two cases: when the monodromy map of the compatible open book is periodic and when it is pseudo-Anosov. For an open book with periodic monodromy, we verify the Weinstein conjec…

The modified Cholesky decomposition is commonly used for precision matrix estimation given a specified order of random variables. However, the order of variables is often not available or cannot be pre-determined. In this work, we propose to address the variable order issue in the modified Cholesky decomposition for sp…

Many machine learning applications use latent variable models to explain structure in data, whereby visible variables (= coordinates of the given datapoint) are explained as a probabilistic function of some hidden variables. Finding parameters with the maximum likelihood is NP-hard even in very simple settings. In rece…

We study the structure of finite quandles in terms of subquandles. Every finite quandle $Q$ decomposes in a natural way as a union of disjoint $Q$-complemented subquandles; this decomposition coincides with the usual orbit decomposition of $Q$. Conversely, the structure of a finite quandle with a given orbit decomposit…