Study post-hoc Learning to Defer using density-ratio losses.
problem Optimizing decision-making between models and experts.
method Density-ratio losses for post-hoc L2D scorers, derived from class-probability estimation.
result The approach recovers known results and introduces new connections to expert comparison and anomaly detection.
Framework explains deep learning generalization by comparing real and ideal worlds.
problem Understanding why deep models generalize well in practice.
method Integrates real-world empirical loss with ideal population loss to decompose test error.
result The gap between real and ideal worlds is small in deep learning, suggesting robust optimization leads to good generalization.
Curriculum Learning - the idea of teaching by gradually exposing the learner to examples in a meaningful order, from easy to hard, has been investigated in the context of machine learning long ago. Although methods based on this concept have been empirically shown to improve performance of several learning algorithms, …
Equations track profits and losses in trading algorithms.
problem Evaluating the performance of trading algorithms.
method Formal equations incorporating spread.
result Evaluate trading model algorithms' performance.
Paper proposes a method to estimate counterfactual outcomes without a known SCM.
problem Estimating counterfactual outcomes without a known structural causal model.
method Introduces rank preservation assumption and a novel ideal loss for unbiased learning of counterfactual outcomes.
result The proposed method is effective and unbiased, as shown by theoretical analysis and experiments.
A new approach to model rejection using density ratios.
problem Improving model performance through selective prediction.
method Optimization of a loss's risk with φ-divergence regularization to find an idealized data distribution.
result Model rejection can be made by comparing the density ratio of the idealized distribution to the actual data distribution.
This paper tackles selection bias in recommender systems by considering the neighborhood effect.
problem Selection bias in recommender systems due to filtering and user selection.
method Formalizes neighborhood effect as interference problem, introduces treatment representation, and proposes ideal loss.
result Proposed methods achieve unbiased learning when both selection bias and neighborhood effect are present.
We present α-loss, α∈[1,∞], a tunable loss function for binary classification that bridges log-loss (α=1) and 0-1 loss (α=∞). We prove that α-loss has an equivalent margin-based form and is classification-calibrated, two desirable properties for a good surrogate loss function for the ideal y…
We consider the problem of learning a loss function which, when minimized over a training dataset, yields a model that approximately minimizes a validation error metric. Though learning an optimal loss function is NP-hard, we present an anytime algorithm that is asymptotically optimal in the worst case, and is provably…
This paper analyzes and compares different Automated Market Maker mechanisms.
problem Impermanent loss in Constant Function Market Makers.
method Mean-Variance analysis of liquidity providers' profit and loss, comparison of different mechanisms.
result Optimized oracle-based mechanisms outperform Constant Function Market Makers.
A new learning method using hyperbolic geometry for class labels.
problem Class label representation and learning in machine learning.
method Hyperbolic Prototype Learning with a new loss function based on hyperbolic geometry.
result Hyperbolic Prototype Learning is equivalent to logistic regression in the one-dimensional case.
Develops gradient boosting for multi-label classification.
problem Lack of customizable learning algorithms for multi-label classification.
method Generalizes gradient boosting to multi-output problems and proposes an algorithm for learning multi-label classification rules.
result Ability to minimize both decomposable and non-decomposable loss functions.
Active learning is a type of sequential design for supervised machine learning, in which the learning algorithm sequentially requests the labels of selected instances from a large pool of unlabeled data points. The objective is to produce a classifier of relatively low risk, as measured under the 0-1 loss, ideally usin…
We present GLASSES: Global optimisation with Look-Ahead through Stochastic Simulation and Expected-loss Search. The majority of global optimisation approaches in use are myopic, in only considering the impact of the next function value; the non-myopic approaches that do exist are able to consider only a handful of futu…
Paper connects rejection learning to Bhattacharyya divergence.
problem Learning models to abstain from predictions.
method Developed a link between rejection and thresholding different statistical divergences, focusing on Bhattacharyya divergence.
result Rejector obtained by joint ideal distribution corresponds to thresholding of skewed Bhattacharyya divergence.
Focal loss improves deep neural networks' accuracy and calibration.
problem Miscalibration in deep neural networks.
method Using focal loss and temperature scaling to improve model calibration.
result Focal loss leads to state-of-the-art calibrated models without sacrificing accuracy.
Adversarial training achieves optimal test error for shallow networks.
problem Achieving optimal adversarial test error for general data distributions.
method Applying new Rademacher complexity bounds and properties of optimal adversarial predictors.
result Adversarial training can achieve optimal adversarial test error for general data distributions.
Toolbox for stochastic Euler equations using Ebin-Marsden theory.
problem Applying geometric framework to stochastic PDEs.
method Combining infinite-dimensional geometry and stochastic analysis.
result Local well-posedness of maximal solutions for incompressible Euler equation with noise.
This paper proposes a Convolutional Neural Network (CNN) inspired by Multitask Learning (MTL) and based on speech features trained under the joint supervision of softmax loss and center loss, a powerful metric learning strategy, for the recognition of emotion in speech. Speech features such as Spectrograms and Mel-freq…
The study proves poor ideal three-edge triangulations are minimal for certain 3-manifolds.
problem Finding minimal ideal triangulations for specific 3-manifolds.
method Analyzing properties of poor ideal three-edge triangulations and applying them to construct minimal triangulations.
result Poor ideal three-edge triangulations are proven to be minimal for certain 3-manifolds.
Super-acceleration of gradient descent with momentum improves loss function minimization.
problem Minimizing loss functions in machine learning.
method Extending Nesterov acceleration by using gradients at multiple steps ahead.
result Super-acceleration of the momentum algorithm is beneficial for various loss landscapes and tasks.
We give a simple method to find ideal points of the character variety of a 3-manifold from an ideal triangulation.
Recommender systems widely use implicit feedback such as click data because of its general availability. Although the presence of clicks signals the users' preference to some extent, the lack of such clicks does not necessarily indicate a negative response from the users, as it is possible that the users were not expos…
Proposes measures for uncertainty quantification using proper scoring rules.
problem Uncertainty quantification for prediction tasks.
method Decomposes proper scoring rules into divergence and entropy components, tailoring uncertainty quantification to specific tasks.
result Flexibility in uncertainty quantification improves performance in selective prediction and active learning.
We formalize how markets aggregate via arbitrage and quantify liquidity loss.
problem How financial markets aggregate and the loss of liquidity.
method Characterize markets via utility functions, use thermodynamics analogy, derive limit order book representation, compute aggregation loss.
result Arbitrage-mediated aggregation leads to market-dynamical entropy quantifying liquidity loss.
Sparse model selection by structural risk minimization leads to a set of a few predictors, ideally a subset of the true predictors. This selection clearly depends on the underlying loss function L~. For linear regression with square loss, the particular (functional) Gradient Boosting variant L2−Boosting exce…
Paper provides new Alexander ideal-based obstruction to 0-concordance of knotted surfaces.
problem Tackles the 0-concordance problem for knotted surfaces in S4. method Uses Alexander ideals to induce a homomorphism and prove non-sliceness.
result Alexander ideal determines 0-concordance classes and non-sliceness.
Paper finds sparse representation of functions using inverse scale space flow.
problem Finding sparse representation of L2 functions. method Inverse scale space flow to minimize L2 loss. result Convergence to optimal solution in ideal and noisy cases.
Short proof for ideal polygons with near optimal orthogeodesic decomposition.
problem Decomposing ideal polygons into orthogeodesics.
method Short proof with orthogeodesic decomposition of length at most 2log(n). result Optimal orthogeodesic decomposition of ideal polygons with length 2log(n). This work simplifies adversarial attacks using neural networks, reducing computation and improving training convergence.
problem Efficiently generating and training against ideal adversarial attacks with minimal computational overhead.
method Representing ideal adversarial attacks as smooth piece-wise functions and approximating them with neural networks. Using a mathematical game between an attack network and a defense network for adversarial training.
result Obtained convergence rates of adversarial loss in terms of sample size n for adversarial training. A method uses CG to create efficient channels for ideal observers.
problem Computational intractability of ideal observers for high-dimensional image data.
method Conjugate gradient (CG) method for constructing efficient channels.
result CG-based channels approximate IO and HO performance efficiently.
The goal of this work is to study the ideals of the Goldman Lie algebra S. To do so, we construct an algebra homomorphism from S to a simpler algebraic structure, and focus on finding ideals of this new structure instead. The structure S can be regarded as either a Q-module or a Q-module gen…
Study introduces dynamical ideals for non-commutative rings and classifies knots and links.
problem Classifying surface knots and links in smooth 4-manifolds.
method Introduced dynamical analog of prime ideals for non-commutative rings and proved a factorization theorem.
result Classified surface knots and links in smooth 4-manifolds.
We investigate the rigidity of hyperbolic cone metrics on 3-manifolds which are isometric gluing of ideal and hyper-ideal tetrahedra in hyperbolic spaces. These metrics will be called ideal and hyper-ideal hyperbolic polyhedral metrics. It is shown that a hyper-ideal hyperbolic polyhedral metric is determined up to i…
Defines timelike ideal boundary for non-positively curved Lorentzian spaces.
problem Understanding the geometry of non-positively curved Lorentzian spaces.
method Introduces timelike ideal boundary as asymptotic classes of geodesic rays, endows with topology and metric, and studies upper curvature bounds.
result Established upper curvature bounds for the resulting metric space.
A taut ideal triangulation of a 3-manifold is a topological ideal triangulation with extra combinatorial structure: a choice of transverse orientation on each ideal 2-simplex, satisfying two simple conditions. The aim of this paper is to demonstrate that taut ideal triangulations are very common, and that their behavio…
Second-order optimizers retain residual information after data deletion, affecting machine unlearning.
problem Residual information in second-order optimizers after data deletion.
method Comparison of first-order and second-order learners, eigendecomposition analysis.
result Second-order optimizers retain residual information, not detectable by first-order analysis.
The paper studies dynamical properties in semigroups modulo ideals.
problem Analyzing shadowing, expansivity, and stability in semigroups with ideals.
method Investigates shadowing, expansivity, and stability properties in uniform transformation semigroups modulo an ideal.
result Establishes that if a semigroup exhibits shadowing and expansivity modulo an ideal, it is also topologically stable modulo that ideal.
We define the ideal simplicial volume for compact manifolds with boundary. Roughly speaking, the ideal simplicial volume of a manifold M measures the minimal size of possibly ideal triangulations of M "with real coefficients", thus providing a variation of the ordinary simplicial volume defined by Gromov in 1982, t…
The notion of ideal immersions was introduced by the author in 1990s. Roughly speaking, an ideal immersion of a Riemannian manifold into a real space form is a nice isometric immersion which produces the least possible amount of tension from the ambient space at each point. In this paper, we classify all ideal hypersur…
New formula calculates volumes of ideal hyperbolic drums.
problem Computing volumes of ideal hyperbolic drums.
method Proved a volume formula for arbitrary ideal hyperbolic antiprisms (drums).
result Volume formula for ideal hyperbolic drums.
Study of combinatorial Calabi flow on ideal circle patterns.
problem Finding ideal circle patterns with prescribed curvatures.
method Combinatorial Calabi flow in hyperbolic and Euclidean geometry.
result Flow converges exponentially to ideal circle patterns.
S2M optimizes mining for diverse data subpopulations.
problem Scalability and uniformity in training sets with many labels and diverse data.
method Doubly-stochastic mining (S2M) computes per-example and minibatch losses on hardest labels/examples.
result S2M ensures good performance across all data subpopulations.
A biconservative submanifold of a Riemannian manifold is a sub- manifold with divergence free stress-energy tensor with respect to bienergy. These are generalizations of biharamonic submanifolds. In 2013, B. Y. Chen and M.I. Munteanu proved that δ(2)-ideal and δ(3)-ideal biharmonic hypersurfaces in Euclidean space …
The paper studies deformations of Lie ideals in Lie algebras.
problem Understanding deformations of Lie ideals in Lie algebras.
method Develops deformation theory, compares cohomologies, enriches deformation complex.
result Deformation cohomology classes differentiate smooth deformations of ideals.
Paper solves long-standing problem of infinite ideal polyhedra in hyperbolic space.
problem Characterize infinite ideal polyhedra in hyperbolic 3-space.
method Introduced combinatorial Ricci flow for infinite ideal circle patterns.
result Proved characterization of infinite ideal circle patterns under specific conditions.
Wintgen ideal surfaces in E^4 form an important family of surfaces, namely surfaces with circular ellipse of curvature. Obviously, Wintgen ideal surfaces satisfy the pointwise equality K+K_N=H^2. In the present study we consider the Wintgen ideal surfaces in n-dimensional Euclidean space E^4. We have shown that Wintgen…
We discuss two different in general natural approaches to the ideal closure and ideal boundary of Busemann nonpositively curved metric space. It is shown that the identity map of the space admits surjective continuation from its coarse ideal closure to the weak one. We consider some situations when these closures coinc…