A new point process for clustering distributions with repulsion.
problem Clustering distributions with repulsion.
method Distributional Determinantal Point Process (dDPP) with sliced Wasserstein kernel.
result Validated dDPP as a well-defined point process and applied to gene expression and epilepsy data.
Extends uncertainty detection in neural networks to finer distinctions.
problem Detecting finer distinctions between certain, uncertain, and out-of-distribution points.
method Two-step approach: first builds class distribution using Kernel Activation Vectors, second determines test point confidence.
result Corrects overconfident NN decisions and learns to say 'I don't know' when uncertain.
A new framework evaluates model performance on single input points, revealing insights into data and model structure.
problem Traditional evaluation methods in machine learning are insufficient for understanding model performance and data structure.
method Developed a pointwise framework to measure model performance on individual input points, analyzing the relationship between pointwise and average performance.
result Profiles of data points reveal different types of correlations between pointwise and average performance, challenging existing models of learning.
DDEQs extend DEQs to discrete measure inputs using Wasserstein gradient flows.
problem Applying DEQs to discrete measure inputs like sets or point clouds.
method Wasserstein gradient flows for finding fixed points of discrete measures under permutation-invariance.
result DDEQs can compete with state-of-the-art models in tasks like point cloud classification and completion.
Distributions over exchangeable matrices with infinitely many columns, such as the Indian buffet process, are useful in constructing nonparametric latent variable models. However, the distribution implied by such models over the number of features exhibited by each data point may be poorly- suited for many modeling tas…
Proposes a new method to model event sequences using normalizing flows.
problem Modeling asynchronous and probabilistic event sequences.
method Intensity-free framework using normalizing flows.
result Effective at capturing stochasticity of discrete event sequences.
Estimates change point in high-dimensional dynamic graphical models.
problem Detecting change points in high-dimensional graphical models.
method Developed an estimator with Op(ψ−2) rate of convergence, established asymptotic distribution under high-dimensional scaling. result Asymptotic distribution characterized under vanishing and non-vanishing jump size regimes.
The paper develops methods to accurately locate change points in high-dimensional mean shift models.
problem Locating change points in high-dimensional mean shift models.
method Locally refitted least squares estimator, component-wise and simultaneous rates of estimation.
result Asymptotic validity of component-wise and simultaneous confidence intervals for change point parameters.
Model change points in time-series data with neural SDEs and variational autoencoders.
problem Modeling change points in time-series data with neural stochastic differential equations.
method Proposes a novel model formulation and training procedure based on the variational autoencoder framework, alternating between updating neural SDE parameters and change points.
result Demonstrates the expressive power of the proposed model in modeling both classical parametric SDEs and real datasets with distribution shifts.
We present a homogenization theorem for isotropically-distributed point defects, by considering a sequence of manifolds with increasingly dense point defects. The loci of the defects are chosen randomly according to a weighted Poisson point process, making it a continuous version of the first passage percolation model.…
Improved reasoning model by sampling from power distribution without additional training.
problem Efficiently sampling from a sharpened distribution to improve reasoning models.
method Entropy-Cut Metropolis-Hastings algorithm that identifies key decision points for resampling.
result The method consistently improves reasoning models across various datasets.
ELBO converges to a sum of entropies for many generative models.
problem Understanding the convergence of variational lower bounds in unsupervised learning.
method Analyzing the ELBO for a broad class of generative models, showing it equals a sum of entropies.
result The ELBO is equal to a sum of entropies at stationary points for many generative models.
EFDM models spatial point processes with variable cardinality using existence variables.
problem Challenges in extending diffusion models to variable-cardinality spatial point processes.
method Existence-field diffusion model (EFDM) that jointly models spatial locations and cardinality without discrete transitions.
result EFDM achieves improved modeling capability on datasets with varying cardinality.
A new framework assigns values to data points considering their distribution.
problem Limited applicability of data Shapley to points outside the fixed data set.
method Proposes distributional Shapley, defining point value in context of data distribution.
result Distributional Shapley values are stable under data point and distribution perturbations.
PerfGD solves model-induced data shifts by finding optimal points.
problem Model-induced data shifts where model choice changes data distribution.
method Performative Gradient Descent (PerfGD) which explicitly captures model-data interactions.
result PerfGD provably converges to performatively optimal point.
ManiFlow models manifold data by optimizing NFs on perturbed data.
problem Capturing manifold data with NFs' invertibility constraint.
method Train NFs on perturbed data to implicitly represent manifold.
result NFs implicitly model manifold in regions of maximum likelihood.
New method upsamples sparse, non-uniform point clouds more accurately.
problem Suboptimal results from existing point cloud upsampling methods.
method Imposes manifold distribution constraints using Gaussian functions.
result Generates higher-quality, more uniformly distributed dense point clouds.
EBPs model exchangeable data with flexible distributions.
problem Current energy-based models restrict set cardinality and limited distribution forms.
method Introduced Energy-Based Processes (EBPs) that extend energy models to exchangeable data with neural network parameterizations.
result EBPs can express more flexible distributions over sets without cardinality restrictions.
Detects change points in time series focusing on specific components.
problem Identifying moments when specific components of multivariate time series change distributions.
method Two-stage non-parametric algorithm: causal structure learning followed by change point detection.
result Validated the approach on synthetic and real-world datasets.
3D point cloud attacks examine how neural networks can be fooled.
problem Understanding how 3D neural networks can be exploited by attackers.
method Examined two categories of attacks: distributional and shape attacks.
result Some shape attacks can fool 3D point cloud classification models even after preprocessing.
Paper tests DPPs for diversity models, distinguishing them from other distributions.
problem Testing whether a given distribution is a Determinantal Point Process (DPP) or far from any DPP.
method Proposes the first algorithm for DPP testing and establishes a lower bound on sample complexity.
result Establishes a matching lower bound on the sample complexity of DPP testing.
A method for fast estimation of Wasserstein distances using sliced Wasserstein distances.
problem Efficiently computing Wasserstein distances for multiple pairs of distributions.
method Regression on sliced Wasserstein distances to predict true Wasserstein distances.
result The proposed method provides a better approximation of Wasserstein distance than state-of-the-art models, especially in low-data regimes.
The paper provides convergence bounds for approximating a distribution using point clouds.
problem Approximating a distribution using discrete points with minimal Wasserstein distance.
method Lloyd's algorithm with Power cells, analyzed using gradient descent.
result Explicit upper bounds for the convergence speed of the Lloyd-type algorithm.
New methods connect low-loss points on neural network surfaces.
problem Connecting low-loss points on neural network loss surfaces.
method Macroscopic distributional assumptions and global connection models.
result Accuracy correlates with complexity and sensitivity.
Bayesian method adapts to unknown distribution shifts in online learning.
problem Online learning with unknown and irregular distribution shifts.
method Bayesian inference with change-point detection and beam search.
result Improves adaptation to new data distributions over state-of-the-art methods.
We consider the problem of quickest change-point detection in data streams. Classical change-point detection procedures, such as CUSUM, Shiryaev-Roberts and Posterior Probability statistics, are optimal only if the change-point model is known, which is an unrealistic assumption in typical applied problems. Instead we p…
Extends DAMs to Gaussian distributions for efficient pattern storage and retrieval.
problem Limited storage capacity and retrieval methods for non-vector pattern representations.
method Introduces a log-sum-exp energy function over Gaussian distributions, using optimal transport maps for retrieval dynamics.
result Proves exponential storage capacity and provides quantitative retrieval guarantees.
Extends Hawkes process for flexible residual modeling in point processes.
problem Modeling high-frequency financial data with complex residual distributions.
method Introduces self and mutually exciting point process with discretely Markovian dynamics.
result Flexible residual distributions improve intensity modeling and high-frequency data estimation.
We study the Immediate Exchange model, recently introduced by Heinsalu and Patriarca [Eur. Phys. J. B 87: 170 (2014)], who showed by simulations that the wealth distribution in this model converges to a Gamma distribution with shape parameter 2. Here we justify this conclusion analytically, in the infinite-population…
The paper addresses data uncertainty in graph embedding by modeling data points as Gaussian distributions.
problem Data uncertainty in machine learning pipelines leads to misleading embeddings and lower accuracy.
method The paper proposes modeling data uncertainty using Gaussian distributions and reformulates graph embedding techniques.
result The proposed methods improve the accuracy of graph embedding by accounting for data uncertainty.
This paper calibrates distribution models from PELVE values.
problem Calibrating distribution models to match given PELVE values.
method Discusses various calibration methods for PELVE under different constraints.
result Developed techniques to convert PELVE calibration to advanced differential equations.
A new clustering method preserves data distribution.
problem Distorted cluster centers in k-means clustering.
method Distributional Clustering method ensuring cluster centers mimic data distribution.
result Cluster centers converge to data generating distribution.
New method improves probabilistic electricity price predictions.
problem Improving point forecasts to probabilistic distributions for better decision-making.
method Isotonic Distributional Regression combined with other postprocessing methods.
result Isotonic Distributional Regression outperforms other methods in combining probabilistic distributions.
Proposes neural SDEs with change points for better time series modeling.
problem Restrictions in modeling time series with distributional shift.
method Generative adversarial networks (GANs) for SDEs and change point detection.
result Jointly learns change points and SDE model parameters.
Proposes a framework for modeling RTB auctions using point processes.
problem Modeling and optimizing repeated auctions in the RTB ecosystem.
method Develops a stochastic framework using point processes to model and optimize RTB auctions.
result The proposed framework can be approximated to a Poisson point process, enabling the use of established properties.
Estimates change point in dynamic stochastic block model.
problem Estimating the location of a single change point in a dynamic stochastic block model.
method Two methods: least squares with clustering and ignoring community structures.
result Established rates of convergence and asymptotic distributions of change point estimators.
Given a heterogeneous time-series sample, the objective is to find points in time (called change points) where the probability distribution generating the data has changed. The data are assumed to have been generated by arbitrary unknown stationary ergodic distributions. No modelling, independence or mixing assumptions…
Pairs (Hamiltonian system, Lagrangian distribution), called dynamical Lagrangian distributions, appear naturally in Differential Geometry, Calculus of Variations and Rational Mechanics. The basic differential invariants of a dynamical Lagrangian distribution w.r.t. the action of the group of symplectomorphisms of the a…
Improves point-cloud reconstruction by optimizing projections with self-attention.
problem Inefficient and non-metric projection methods for sliced Wasserstein distances.
method Proposes distributional sliced Wasserstein distance with self-attention for permutation-invariant and metric optimization.
result Self-attention amortized distributional projection optimization achieves better performance in point-cloud reconstruction.
Causal influence measures for machine learnt classifiers shed light on the reasons behind classification, and aid in identifying influential input features and revealing their biases. However, such analyses involve evaluating the classifier using datapoints that may be atypical of its training distribution. Standard me…
A new method models galaxies as points in space for better analysis.
problem Limitations of binning and voxelization in galaxy surveys.
method A diffusion-based generative model for galaxy point clouds.
result Demonstrated on dark matter haloes in Quijote simulations.
A new method speeds up Gaussian Process inference for large datasets.
problem Expensive inference for GP models with non-Gaussian noises.
method Construct a variational distribution considering only short-distance correlations, using Graph Convolutional Networks for model reuse.
result Significantly speeds up inference and often gets more accurate results.
Bayesian meta-reinforcement learning improves over point estimates with Laplace approximation.
problem Improving meta-reinforcement learning by providing full posterior distributions.
method Augmenting point estimates with Laplace approximation for full posterior distributions.
result Our method performs similarly to variational baselines with fewer parameters.
Tukey median performance analyzed under TV corruptions.
problem Performance analysis of Tukey median under TV corruptions.
method Analysis of Tukey median and projection algorithm under TV corruptions.
result Breakdown point reduced to 1/4 under TV corruptions, compared to 1/3 under Huber's model.
DSPPs improve predictive distributions in scalable regression tasks.
problem Improving predictive distributions in scalable regression tasks.
method Inspired by DGPs, DSPPs use mini-batch training and kernel basis functions for uncertainty control.
result DSPPs provide significantly better calibrated predictive distributions than other methods.
New model learns function distributions from datasets.
problem Learning function distributions from datasets.
method Functional Neural Processes (FNPs) model distributions over functions by learning a graph of dependencies on top of latent representations.
result FNPs offer competitive predictions and more robust uncertainty estimates compared to baselines.
Generative models learn distributions of continuous functions.
problem Training generative models on discretized grids limits model size and data type.
method Parameterize data points by continuous functions, learn distributions over these functions.
result Models can learn rich distributions of functions independently of data type and resolution.
New model captures time and mark inter-dependence in TPPs.
problem Limited predictive performance of conditionally independent TPP models on entangled time and mark interactions.
method Developed a multivariate TPP that models conditional inter-dependence of time and mark, using both intensity-based and intensity-free models.
result Proposed TPP models outperform conditionally independent and dependent models in standard prediction tasks.