arXiv research

Study on supply chain networks using wire transfers in Brazil.

Transfer learning is a popular practice in deep neural networks, but fine-tuning of large number of parameters is a hard task due to the complex wiring of neurons between splitting layers and imbalance distributions of data in pretrained and transferred domains. The reconstruction of the original wiring for the target …

Differentiable NAS frameworks grow networks wider and deeper, revealing biases in wiring evolution.

(NOTE: per referee comments, this article has been split; it is now superseded by "Existence of thread-wire minimizers" and "Near-wire thread-wire minimizers"; please see http://www.bkstephens.net.) Alt's thread problem asks for least-area surfaces bounding a fixed "wire" curve and a movable "thread" curve of length L.…

Wire billiard is defined by a smooth embedded closed curve of non-vanishing curvature $k$ in $\mathbb{R}^n$ (a wire). For a class of curves, that we call nice wires, the wire billiard map is area preserving twist map of the cylinder. In this paper we are investigating whether the basic features of conventional planar b…

Decussation improves robustness of 3D neural networks.

There are several types of equation of motion of elastic wires. In this paper, we treat an equation taking account of the thickness of wire. The equation was introduced by Caflisch and Maddocks on plane curves, and they proved the existence of solutions. Koiso and Sugimoto generalized the result to any dimensional Eucl…

Theoretical comparison of three invariance approaches in deep linear networks.

Researchers use Gaussian Process Regression to improve accuracy of a low-cost hot-wire anemometer.

Generative AI agents improve ERP systems by automating complex financial tasks.

Enneper's wire, the image of the circle of radius $R$ under Enneper's surface, bounds exactly three minimal surfaces for $R$ between 1 and $\sqrt 3$, and these three surfaces depend continuously on $R$. The other two surfaces (besides Enneper's surface) are absolute minima of area among disk-type surfaces bounded by En…

Unbraided wiring diagrams for Stein fillings of lens spaces are described.

RicciNets prunes neural networks by removing edges of low importance based on Ricci curvature, reducing FLOPs by 35%.

Generative model for morphisms in free categories learns from wiring diagrams.

In-vivo examination of the physical connectivity of axonal projections through the white matter of the human brain is made possible by diffusion weighted magnetic resonance imaging (dMRI) Analysis of dMRI commonly considers derived scalar metrics such as fractional anisotrophy as proxies for "white matter integrity," a…

EM-GAN uses GANs for fast stress analysis of multi-segment interconnects.

Serenity optimizes neural network execution for edge devices by scheduling with optimal memory footprint.

We study association between macroeconomic news and stock market returns using the statistical theory of copulas, and a new comprehensive measure of news based on the indexing of news wires. We find the impact of economic news on equity returns to be nonlinear and asymmetric. In particular, controlling for economic con…

Soap films collapse only if their bulk has negative pressure, forming convex shapes.

Estimates for graph embeddings into symmetric spaces derived from coarse geometry.

We give an upper bound on the z-degree of the Kauffman polynomial of a link, using bridges of length greater than one which are separated in some tangle decomposition of a link diagram. We construct some examples by wiring together rational tangles.

What are the possible shapes of various things and why? For instance, when a closed wire or a frame is dipped into a soap solution and is raised up from the solution, the surface spanning the wire is a soap film. What are the possible shapes of soap films and why? Or, for instance, why is DNA like a double spiral stair…

Paper uses CNN to predict process parameters from molten pool data in WLAM.

In mathematics, the classical Plateau problem consists of finding the surface of least area that spans a given rigid boundary curve. A physical realization of the problem is obtained by dipping a stiff wire frame of some given shape in soapy water and then removing it; the shape of the spanning soap film is a solution …

Survey of integrable billiard models and inequalities.

The paper explores knots with equal bridge and braid index, conjecturing they have a unique equilibrium state.

Motivated by the study of the equilibrium equations for a soap film hanging from a wire frame, we prove a compactness theorem for surfaces with asymptotically vanishing mean curvature and fixed or converging boundaries. In particular, we obtain sufficient geometric conditions for the minimal surfaces spanned by a given…

We introduce a new neural network model, together with a tractable and monotone online learning algorithm. Our model describes feed-forward networks for classification, with one output node for each class. The only nonlinear operation is rectification using a ReLU function with a bias. However, there is a rectifier on …

New neural net learns time-reversible symplectic dynamics.

Study on dynamic curves with elastic energy and spontaneous curvature.

Method finds differential equations for integrable billiard tables.

Low-density parity-check codes, a class of capacity-approaching linear codes, are particularly recognized for their efficient decoding scheme. The decoding scheme, known as the sum-product, is an iterative algorithm consisting of passing messages between variable and check nodes of the factor graph. The sum-product alg…

Transfer learning improves portfolio optimization by identifying transfer risk.

Paper analyzes transfer risk in transfer learning for finance.

Current algorithms for deep learning probably cannot run in the brain because they rely on weight transport, where forward-path neurons transmit their synaptic weights to a feedback path, in a way that is likely impossible biologically. An algorithm called feedback alignment achieves deep learning without weight transp…

This paper explores the connection between adversarial and knowledge transferability.

Mathematical framework for transfer learning feasibility and transfer risk.

Study on evolving interfaces with complex curvature and density effects.

Transfer learning borrows knowledge from a source domain to facilitate learning in a target domain. Two primary issues to be addressed in transfer learning are what and how to transfer. For a pair of domains, adopting different transfer learning algorithms results in different knowledge transferred between them. To dis…

Bayesian method reconstructs neural network memories from connectivity.

The introduction of automated flight control and management systems have made possible aircraft designs that sacrifice arodynamic stability in order to incorporate stealth technology intro their shape, operate more efficiently, and are highly maneuverable. Therefore, modern flight management systems are reliant on mult…

Many transformations in deep learning architectures are sparsely connected. When such transformations cannot be designed by hand, they can be learned, even through plain backpropagation, for instance in attention mechanisms. However, during learning, such sparse structures are often represented in a dense form, as we d…

Study measures impact of data and neural net similarity on transferability in restaurant sales data.

The paper analyzes phase transitions in transfer learning for perceptrons.

We investigate the elastic behavior of knotted loops of springy wire. To this end we minimize the classic bending energy $E_{\text{bend}}=\intκ^2$ together with a small multiple of ropelength $\mathcal R=\text{length}/\text{thickness}$ in order to penalize selfintersection. Our main objective is to characterize elastic…

Adaptive source selection for positive transfer in linear models improves target dataset performance.

New research on limits of transfer learning, proving key selection and dependence requirements.

Transfer learning aims at improving the performance of target learners on target domains by transferring the knowledge contained in different but related source domains. In this way, the dependence on a large number of target domain data can be reduced for constructing target learners. Due to the wide application prosp…