arXiv research

New algorithms minimize dynamic regret for strongly convex losses.

Paper analyzes proper losses and their performance in machine learning tasks.

The problem of bipartite ranking, where instances are labeled positive or negative and the goal is to learn a scoring function that minimizes the probability of mis-ranking a pair of positive and negative instances (or equivalently, that maximizes the area under the ROC curve), has been widely studied in recent years. …

Let G/H be a strongly regular homogeneous space such that H is a Lie group of inner type. We show that G/H admits a proper action of a discrete non-virtually abelian subgroup of G if and only if G/H admits a proper action of a subgroup L of G locally isomorphic to SL(2,R). We classify all such spaces.

Paper develops proper, lower-bounded losses for weakly supervised classification.

We study proper losses for discrete generative models without knowing the target distribution.

We study losses for binary classification and class probability estimation and extend the understanding of them from margin losses to general composite losses which are the composition of a proper loss with a link function. We characterise when margin losses can be proper composite losses, explicitly show how to determ…

This work broadens calibeating to various proper losses using Bregman divergence.

This work generalizes calibeating for a broader range of proper losses using Bregman divergence.

Optimizing proper loss yields calibrated models under specific conditions.

Study on proper learning under relaxed worst-case robust loss for VC classes.

In this paper, we study strongly Gauduchon metrics on compact complex manifolds. We study the cohomology cones SG in the de Rham cohomology groups generated by all strongly Gauduchon metrics and its direct images under proper modifications. We also study the moduli of strongly Gauduchon manifolds. We prove an existence…

A new method learns proper multiclass losses and probabilities.

Proper learning is possible with labeled data, but unlabeled data can improve performance.

Optimal multiclass U-calibration error found to be Θ(√KT).

The paper explores proper actions and their relation to representation theory, with new quantitative methods.

New algorithm reduces prediction errors across various loss functions.

The goal of online prediction with expert advice is to find a decision strategy which will perform almost as well as the best expert in a given pool of experts, on any sequence of outcomes. This problem has been widely studied and $O(\sqrt{T})$ and $O(\log{T})$ regret bounds can be achieved for convex losses (\cite{zin…

The purpose of this paper is to give a sufficient condition for (strong) stability of non-proper smooth functions (with respect to the Whitney $C^\infty$-topology). We show that a Morse function is stable if it is end-trivial at any point in its discriminant, where end-triviality (which is also called local triviality …

The paper decomposes probabilistic scores into reliability, uncertainty, and information loss.

SA algorithms control dynamic regret in non-stationary settings with strong convexity or exp-concavity.

Proposes measures for uncertainty quantification using proper scoring rules.

Paper proposes fitting loss functions to data using source functions from information geometry.

Strongly polynomial algorithm for approximate Forster transforms and halfspace learning.

The paper disproves the properness conjecture for higher-dimensional minimal hypersurfaces.

In large-scale distributed learning, security issues have become increasingly important. Particularly in a decentralized environment, some computing units may behave abnormally, or even exhibit Byzantine failures -- arbitrary and potentially adversarial behavior. In this paper, we develop distributed learning algorithm…

Gradient EM converges exponentially to optimal solution in agnostic mixtures.

Improves generative models for cost-sensitive decisions.

New method improves decision-making accuracy without complex calculations.

Generative Cross-Entropy improves classification with fewer labels.

Researchers prove the automorphism group of a sphere complex matches the mapping group of a graph.

We consider composite loss functions for multiclass prediction comprising a proper (i.e., Fisher-consistent) loss over probability distributions and an inverse link function. We establish conditions for their (strong) convexity and explore the implications. We also show how the separation of concerns afforded by using …

Optimizes predictions by recalibrating online forecasts with minimal error.

Let $1\to (K,K_1)\to (G,N_G(K_1))\to(Q,Q_1)\to 1$ be a short exact sequence of pairs of finitely generated groups with $K$ strongly hyperbolic relative to proper subgroup $K_1$. Assuming that for all $g\in G$ there exists $k\in K$ such that $gK_1g^{-1}=kK_1k^{-1}$, we prove that there exists a quasi-isometric section $…

This work examines uncertainty sampling in binary classification using equivalent loss.

Embeds CR manifolds into complex spaces using equivariant actions.

Improved online learning with time-varying constraints for complex domains.

The study of a machine learning problem is in many ways is difficult to separate from the study of the loss function being used. One avenue of inquiry has been to look at these loss functions in terms of their properties as scoring rules via the proper-composite representation, in which predictions are mapped to probab…

A new method decomposes subjective risk into epistemic and aleatoric uncertainties.

Estimates proper calibration errors and refinement terms in probabilistic predictions.

Study on privacy leakage in noisy gradient descent algorithms.

The paper explores additional structures on Morse boundaries to distinguish hyperbolic spaces up to quasi-isometry.

New algorithm AG-OG optimizes separable convex-concave problems efficiently.

We study strictly proper scoring rules in the Reproducing Kernel Hilbert Space. We propose a general Kernel Scoring rule and associated Kernel Divergence. We consider conditions under which the Kernel Score is strictly proper. We then demonstrate that the Kernel Score includes the Maximum Mean Discrepancy as a special …

The paper characterizes links in 3D from divides with cusps.

We analyze impermanent loss in AMMs and show G3Ms are simplest.

The paper argues that uncertainty quantification in ML is application-specific and proposes a flexible family of measures.

Improved privacy-preserving methods for convex optimization with heavy-tailed data.