Optimal transport learns Riemannian metrics for evolving probability measures.
problem Learning metrics for evolving probability measures on Riemannian manifolds.
method Neural parametrization of a metric tensor via optimal transport, alternating optimization scheme.
result Improved trajectory inference on scRNA and bird migration data.
Quantum probability metrics improve distribution comparison in high dimensions.
problem Challenges in comparing probability distributions, especially in high-dimensional and non-compact domains.
method Quantum probability metrics (QPMs) derived from quantum state spaces, overcoming limitations of MMD.
result QPMs offer enhanced sensitivity to subtle distributional differences in high dimensions and improve performance in generative modeling.
A new metric for comparing probability measures on graphs, scalable and negative definite.
problem Optimal transport's high complexity and indefiniteness for kernel machines.
method Sobolev transport metric for graph metrics, closed-form formula, negative definiteness.
result Sobolev transport yields a scalable and negative definite metric.
The paper explores geometry of probability measures and barycenter maps.
problem Understanding the space of probability measures and their barycenter.
method Information geometry, Fisher metric, dualistic structures, divergences, geodesics.
result Recent developments in the geometry of probability measures and barycenter.
Permutation invariant network learns Wasserstein metrics.
problem Understanding the space of probability measures and comparing distributions.
method Permutation invariant network mapping samples to a low-dimensional space.
result Network can generalize to compute distances between unseen densities and learn moments.
We show a general relation between the spatially disjoint product of probability density functions and the sum of their Fisher information metric tensors. We then utilise this result to give a method for constructing the probability density functions for an arbitrary Riemannian Fisher information metric tensor. We note…
Paper introduces GSPMs for robust probability metrics.
problem Lack of well-established convergence behavior for probability metrics.
method Introduces Generalized Sliced Probability Metrics (GSPMs) based on generalized Radon transform.
result GSPMs converge to global optimum under mild assumptions for generative modeling.
This work introduces a new metric for comparing imprecise probability models.
problem Quantifying differences between imprecise probability models.
method Integral imprecise probability metric framework based on Choquet integral.
result IIPM enables comparison across different imprecise probability models and quantifies epistemic uncertainty.
We study the minimax optimal rate for estimating the Wasserstein-1 metric between two unknown probability measures based on n i.i.d. empirical samples from them. We show that estimating the Wasserstein metric itself between probability measures, is not significantly easier than estimating the probability measures u…
This paper reviews metrics to assess AI model calibration accuracy.
problem AI model probabilities do not always match their true accuracy.
method Comprehensive review of 82 probability calibration metrics.
result Identified 4 classifier families and 1 object detection family of metrics.
New Schwarz Lemma for Bergman metrics in bounded domains.
problem Finding bounds for Bergman metrics in bounded domains.
method Using Cauchy-Schwarz inequality from probability theory.
result Established a new Schwarz Lemma for Bergman metrics.
We formulate the Riemannian calculus of the probability set embedded with L2-Wasserstein metric. This is an initial work of transport information geometry. Our investigation starts with the probability simplex (probability manifold) supported on vertices of a finite graph. The main idea is to embed the probability m…
A new pseudo-metric uses data depth to compare probability distributions.
problem Designing a metric between probability distributions for machine learning applications.
method Extension of univariate quantiles to multivariate spaces, using data depth and Hausdorff distance.
result The pseudo-metric is robust, factorizes translations, and has good behavior under transformations.
New metric for probability measures connects physics and geometry.
problem Developing a new metric for probability measures.
method Transport Hessian metric, formulated dynamical systems.
result Connections to physics equations and mathematical models.
The Wasserstein metric is an important measure of distance between probability distributions, with applications in machine learning, statistics, probability theory, and data analysis. This paper provides upper and lower bounds on statistical minimax rates for the problem of estimating a probability distribution under W…
We study the minimax optimal rates for estimating a range of Integral Probability Metrics (IPMs) between two unknown probability measures, based on n independent samples from them. Curiously, we show that estimating the IPM itself between probability measures, is not significantly easier than estimating the probabili…
New KQEs improve probability metrics without mean function constraints.
problem Improving probability metrics without relying on mean function representations.
method Kernel quantile embeddings (KQEs) to construct new distances.
result KQEs offer a competitive alternative to MMD with near-linear cost.
Central limit theorem for Green metrics on hyperbolic groups.
problem Proving a central limit theorem for Green metrics on hyperbolic groups.
method Proving a central limit theorem for Green metrics on hyperbolic groups using probability measures and ordering elements.
result Proved a central limit theorem for Green metrics on hyperbolic groups.
New Fourier metrics equivalent to Wasserstein distances in image processing.
problem Equivalence of Fourier-based and Wasserstein metrics in imaging problems.
method Extensions of Fourier-based metrics to handle different centers of mass and discrete measures, showing equivalence to Wasserstein distances.
result New Fourier metrics are equivalent to Wasserstein distances with explicit constants, improving runtime in image processing.
We consider three different approaches to define natural Riemannian metrics on polytopes of stochastic matrices. First, we define a natural class of stochastic maps between these polytopes and give a metric characterization of Chentsov type in terms of invariance with respect to these maps. Second, we consider the Fish…
Paper examines stability of Bayesian posterior measures using integral probability metrics.
problem Stability of Bayesian inference in large-scale inverse problems.
method New families of integral probability metrics for likelihood and prior perturbations.
result Constructs new stability results for Bayesian posterior measures.
Study optimizes tree-based models for better alignment of predicted scores and actual probabilities.
problem Traditional calibration metrics fail to align predicted scores with actual probabilities when score distributions deviate from the underlying data.
method Optimizes tree-based models (Random Forest, XGBoost) using Kullback-Leibler (KL) divergence to minimize the difference between predicted and true probability distributions.
result Optimized tree-based models yield superior alignment between predicted scores and actual probabilities without significant performance loss.
The geometric approach to optimal transport and information theory has triggered the interpretation of probability densities as an infinite-dimensional Riemannian manifold. The most studied Riemannian structures are Otto's metric, yielding the L2-Wasserstein distance of optimal mass transport, and the Fisher--Rao me…
Previous studies have used a specific success metric within an algorithmic search framework to prove machine learning impossibility results. However, this specific success metric prevents us from applying these results on other forms of machine learning, e.g. transfer learning. We define decomposable metrics as a categ…
This work improves deep neural network probability estimation methods.
problem Estimating probabilities from high-dimensional data with inherent uncertainty.
method Investigates and compares methods for probability estimation using deep neural networks, proposing a new method that promotes consistent probabilities.
result The new method outperforms existing approaches on most metrics on simulated and real-world data.
A new IPM uses ReLU networks to measure probability discrepancies.
problem Measuring the difference between two probability distributions in high dimensions.
method Proposes a new parametric IPM using ReLU neural networks to optimize and distinguish between distributions.
result The proposed IPM has good convergence rates and can be used as a surrogate for other IPMs.
A new metric based on hitting probabilities for directed graphs and Markov chains.
problem Lack of metrics specifically adapted to asymmetric structure of directed graphs and Markov chains.
method Metric based on hitting probabilities, insensitive to shortest and average walk distances.
result New structural theory of directed graphs and utility for various applications.
New findings show fixed-kernel discriminators are weaker than feature-learning ones.
problem Comparing performance of fixed-kernel and feature-learning discriminators.
method Using function classes F2 and F1, constructing pairs of distributions, and linking IPMs with sliced Wasserstein distances. result Fixed-kernel IPM and SD cannot discriminate certain distributions that feature-learning IPM and SD can.
Develops RES metrics for stable rare-event forecasting evaluation.
problem Challenges in evaluating forecasts of rare events.
method Rare-event-stable (RES) metrics designed to maintain stable thresholds under extreme rarity.
result RES metrics maintain stable thresholds, consistent model rankings, and near-complete prevalence invariance.
The paper explores statistical and topological properties of sliced probability divergences.
problem Understanding the topological, statistical, and computational consequences of slicing divergences.
method Deriving theoretical properties of sliced probability divergences, including metric axioms preservation and weak continuity.
result Sliced divergences share similar topological properties and have stable sample complexity.
NNLMs optimize poorly for word probabilities due to embedding space structure.
problem NNLMs assign suboptimal probabilities to some words.
method Analyzed the inductive bias of NNLMs and the structure of word embeddings.
result Words on the convex hull have bounded probability, affecting others.
New dimension concept for groups based on percolation probability.
problem Defining a new dimension for groups using percolation probability.
method Introducing percolation dimension pdim(G) for groups G using symmetric probability measures. result The percolation dimension pdim(G) has natural properties like monotonicity and coincides with growth rate exponents for various groups. This paper argues against using calibration metrics for assessing posterior probabilities and proposes expected proper scoring rules instead.
problem The assessment of posterior probabilities generated by machine learning classifiers using calibration metrics is flawed and should be replaced with expected proper scoring rules.
method The paper reviews proper scoring rules from a practical perspective, explains why expected PSRs are a principled measure of posterior quality, and introduces a new calibration metric called calibration loss.
result Calibration loss is superior to expected calibration error and expected score divergence calibration metrics for assessing posterior probabilities.
Develops a new divergence framework that combines f-divergences and IPMs.
problem Comparing distributions that are not absolutely continuous.
method Introduces (f,Γ)-divergences as a two-stage mass-redistribution/mass-transport process. result Improves estimation, learning, and uncertainty quantification in GANs for heavy-tailed distributions.
The paper introduces metrics to rank potential outcomes for better decision-making.
problem Optimal action selection in uncertain situations using causal reasoning.
method Introducing two new metrics: probabilities of potential outcome ranking (PoR) and probability of achieving the best potential outcome (PoB). Establishing identification theorems and deriving bounds for these metrics, and presenting estimation methods.
result The estimators' finite-sample properties and their application to a real-world dataset are demonstrated.
New bounds use IPMs to improve generalization in machine learning.
problem Improving generalization bounds in machine learning.
method PAC-Bayes bounds with Integral Probability Metrics (IPM).
result Natural interpolation between worst-case and favorable cases.
A key prerequisite to optimal reasoning under uncertainty in intelligent systems is to start with good class probability estimates. This paper improves on the current best probability estimation trees (Bagged-PETs) and also presents a new ensemble-based algorithm (MOB-ESP). Comparisons are made using several benchmark …
Study shows zero probability of cut locus for Fréchet mean on Riemannian manifolds.
problem Understanding the cut locus of Fréchet mean on Riemannian manifolds.
method Analytical proof and examples.
result Cut locus of Fréchet mean has zero probability.
Identifies most probable flows for Kunita SDEs in fluid dynamics.
problem Modeling stochastic processes with Eulerian noise and deterministic drifts.
method Equipping the domain with a Riemannian metric from the noise, solving the resulting PDEs.
result Most probable flows differ from deterministic flows, especially under noise.
Novel algorithm speeds up computation of Sobolev IPM for graph-based probability measures.
problem Efficient computation of Sobolev IPM for graph-based probability measures.
method Established relation between Sobolev norm and weighted Lp-norm, proposed novel regularization, leveraged graph structure. result Proposed regularized Sobolev IPM provides closed-form expression for fast computation.
Gromov-Wasserstein (GW) is a powerful tool to compare probability measures whose supports are in different metric spaces. GW suffers however from a computational drawback since it requires to solve a complex non-convex quadratic program. We consider in this work a specific family of cost metrics, namely \textit{tree me…
A nonparametric two-sample test using a parametric integral probability metric
problem Detecting distributional differences between two independent samples
method Propose a new two-sample test statistic based on a newly introduced integral probability metric (IPM)
result Establish theoretical guarantees for the associated two-sample testing procedure
The paper provides statistical guarantees for generative models using dimension reduction.
problem Improving the quality of generative models without increasing dimensionality.
method Modeling generative devices as smooth transformations of a lower-dimensional space and using integral probability metrics.
result Established a risk bound showing the impact of dimension reduction on generative model error.
A new distance metric compares probability distributions using kernel covariance operators.
problem Comparing probability distributions in machine learning tasks.
method Introduces a novel distance metric based on Schatten norm of kernel covariance operators.
result The new distance metric is more discriminative and robust to hyperparameters.
Unified view of KL-divergence and IPMs via DRE, with new DRM metrics.
problem Unified understanding of KL-divergence and IPMs.
method Unified representation via maximum likelihood density-ratio estimation (DRE).
result Unified form of IPMs and novel DRM metrics.
The paper proposes a new method for covariate balancing using IPM to improve causal inference.
problem Covariate imbalance in causal inference weighting methods, especially when models are not correctly specified.
method The integral probability metric (IPM) is used to determine optimal weights for treated and control groups.
result The proposed method can be consistent without specifying either the propensity score or outcome regression model.
This research improves demand forecasting by predicting complete probability density functions using machine learning.
problem Forecasting complete probability density functions for better operational decision making.
method Supervised machine learning method 'Cyclic Boosting' for explainable predictions.
result Predicted probability density functions are fully explainable and avoid 'black-box' models.
Random covers of hyperbolic surfaces follow a specific probability measure.
problem Understanding the distribution of random covers of hyperbolic surfaces.
method Analyzing random covers subject to specific group isomorphism conditions.
result Asymptotic distribution of random covers according to a probability measure on moduli space of metric graphs.