arXiv research

Unified feature importance for machine learning models tackles sufficiency and necessity limitations.

Bubbles are essential in certain economic models with high growth and low interest rates.

In 2001, Oestlund conjectured that Reidemeister moves 1 and 3 are sufficient to describe a homotopy from any generic immersion from the circle into the plane to the standard embedding of the circle. We show that this conjecture is false.

Paper establishes a general inequality for warped product CR-submanifolds in Kähler manifolds.

DG algorithms often fail to generalize well in limited domains, highlighting necessary vs. sufficient conditions.

Paper proposes forecast-necessity testing for accurate causal interpretation in nonlinear time-series models.

Paper explores whether gradient normalization can replace clipping for SGD in heavy-tailed noise.

Asymptotic geodesics in convex polygons are convex for large distances.

We give a necessary and sufficient condition for a graph to have a right-angled Artin group as its braid group for braid index $\ge 5$. In order to have the necessity part, graphs are organized into small classes so that one of homological or cohomological characteristics of right-angled Artin groups can be applied. Fi…

Probabilistic models can handle causal inference without special tools.

The paper addresses bias amplification in prediction and decision-making using causal analysis.

Triangulation filters spurious circuits in multilingual models.

Discussing rigidity in codimension 2, extending rigidity concepts.

New framework estimates treatment effects based on preferences.

The present paper deals with the characterization of no-arbitrage properties of a continuous semimartingale. The first main result, Theorem \refMainTheoremCharNA, extends the no-arbitrage criterion by Levental and Skorohod [Ann. Appl. Probab. 5 (1995) 906-925] from diffusion processes to arbitrary continuous semimartin…

Study on multiple linking numbers, extending Gauss diagram formulas.

Characterizes statistical complexity of realizable regression in PAC and online learning.

A neural framework corrects bias in estimating individual treatment effects.

We introduce an up-down coloring of a virtual-link diagram. The colorabilities give a lower bound of the minimum number of Reidemeister moves of type II which are needed between two 2-component virtual-link diagrams. By using the notion of a quandle cocycle invariant, we determine the necessity of Reidemeister moves of…

The paper analyzes GD for KANs, deriving bounds for training, generalization, and privacy.

We show the existence of isoperimetric regions of sufficiently large volumes in general asymptotically hyperbolic three manifolds. Furthermore, we show that large coordinate spheres in compact perturbations of Schwarzschild-anti-deSitter are uniquely isoperimetric. This is relevant in the context of the asymptotically …

We present algorithms for the detection of a class of heart arrhythmias with the goal of eventual adoption by practicing cardiologists. In clinical practice, detection is based on a small number of meaningful features extracted from the heartbeat cycle. However, techniques proposed in the literature use high dimensiona…

The ACS criterion is verified for specific hypersurfaces in unit spheres.

AdaDKRR tackles data silos by combining autonomy, privacy, and collaboration.

While machine learning has proven to be a powerful data-driven solution to many real-life problems, its use in sensitive domains has been limited due to privacy concerns. A popular approach known as **differential privacy** offers provable privacy guarantees, but it is often observed in practice that it could substanti…

Constructs a universal average for Lie algebra elements.

This paper extends Kelly Criterion to include rebalancing frequency for optimal portfolio selection.

Using powerful posterior distributions is a popular approach to achieving better variational inference. However, recent works showed that the aggregated posterior may fail to match unit Gaussian prior, thus learning the prior becomes an alternative way to improve the lower-bound. In this paper, for the first time in th…

Models predict probabilities of causation from limited data.

Paper proposes methods to learn accurate models from incomplete data without imputation.

If $(X,J)$ is an almost complex manifold, then a function $u$ is said to be plurisubharmonic on $X$ if it is upper semi-continuous and its restriction to every local pseudo-holomorphic curve is subharmonic. As in the complex case, it is conjectured that plurisubharmonicity is equivalent to the fact that the $(1,1)$-cur…

Deep nets with massive data learn spatially sparse functions.

We give convergence guarantees for estimating the coefficients of a symmetric mixture of two linear regressions by expectation maximization (EM). In particular, we show that the empirical EM iterates converge to the target parameter vector at the parametric rate, provided the algorithm is initialized in an unbounded co…

Optimal machine learning requires interpolating training data in high-dimensional linear regression.

Using Jeff Holman's comments in Quantitative Finance to illustrate 4 critical errors students should learn to avoid: 1) Mistaking tails (4th moment) for volatility (2nd moment), 2) Missing Jensen's Inequality, 3) Analyzing the hedging wihout the underlying, 4) The necessity of a numeraire in finance.

Sharp analysis improves RLHF sample complexity with KL-regularization.

This study improves user segmentation for online news recommendation systems.

Unified framework for removing unwanted information from machine learning models.

ANT improves TS diffusion models by automatically determining noise schedules.

We prove that every Kaehler solvmanifold has a finite covering whose holomorphic reduction is a principal bundle. An example is given that illustrates the necessity, in general, of passing to a proper covering. We also answer a stronger version of a question posed by Akhiezer for homogeneous spaces of nonsolvable algeb…

Paper characterizes equilibrium strategies for stochastic control with higher-order moments.

New causal analysis reconciles predictive and statistical fairness.

In this work we wish characterize the Einstein manifolds $(M,g)$, however without the necessity of hypothesis of compactness over $M$ and unitary volume of $g$, which are well known in many works. Our result says that if all eingenvalues $λ$ of $r_{g}$, with respect to $g$, satisfy $λ\geq \frac{1}{n}s_{g}$, then $(M,g)…

Deep neural networks map brain lesions to deficits for better brain function understanding.

iSTFTNet speeds up mel-spectrogram vocoders without sacrificing quality.

IGNIS uses neural networks to estimate copula parameters robustly.

Defines Learning Analytics' foundational structure and scope.

Paper tackles covariate shift in offline IL using less proficient behavior data.