Wasserstein
PulseAugur coverage of Wasserstein — every cluster mentioning Wasserstein across labs, papers, and developer communities, ranked by signal.
8 day(s) with sentiment data
-
New semidefinite programming approach for mixture models in machine learning
A new research paper introduces a semidefinite programming approach to approximate target measures using mixtures of distributions, such as Gaussian mixture models. This method is particularly useful for determining mix…
-
New LBDTPP framework generates asynchronous event sequences using latent block diffusion
Researchers have introduced Latent Block-Diffusion Temporal Point Processes (LBDTPP), a new framework designed for generating asynchronous event sequences. This semi-autoregressive approach combines the benefits of auto…
-
New robust Q-learning algorithm tackles mean-field control with Wasserstein uncertainty
Researchers have developed a new robust Q-learning algorithm designed for mean-field control problems. This algorithm addresses challenges posed by Wasserstein uncertainty in common noise laws by integrating a quantizat…
-
New method predicts data distributions under drift and corruption
Researchers have developed a novel online learning method for predicting full data-generating distributions in non-stationary data streams, even when subjected to drift and adversarial corruption. The approach utilizes …
-
TimeLAVA framework offers learning-agnostic data valuation for time series
Researchers have introduced TimeLAVA, a new learning-agnostic framework designed to value temporal segments within time series data. This method addresses limitations of existing approaches by capturing temporal depende…
-
New Theory Guarantees Convergence for Decentralized Diffusion Models
Researchers have established a theoretical convergence guarantee for decentralized diffusion models using ODE-based sampling. This work provides the first Wasserstein-2 distance convergence result for such architectures…
-
New Geometric Framework Unlocks Gaussian Mixture Model Convergence Insights
Researchers have developed a new geometric framework to analyze the convergence rates of parameter estimation in finite Gaussian mixtures. This framework utilizes Hellinger lower bounds to connect density discrepancies …
-
New research advances flow matching models for generative AI
Researchers are exploring advanced techniques for flow matching models, a type of generative model. One paper introduces Gradual Fine-Tuning (GFT), an annealing-based framework to improve stability and efficiency when a…
-
New CROTS framework advances distributional learning evaluation
Researchers have introduced Conditional Random Ordered Transport Spaces (CROTS), a novel framework for evaluating distributional learning. CROTS equips spaces of random probability measures with an ambient Wasserstein m…
-
New papers explore advanced Principal Component Analysis techniques
Two new papers explore advanced Principal Component Analysis (PCA) techniques. One paper, focusing on Wasserstein geometry, introduces a method for analyzing variations in probability distributions using neural networks…
-
Contact Wasserstein Geodesics offer new approach to Schrödinger Bridges
Researchers have developed a novel reformulation of the Schrödinger Bridge problem, termed the non-conservative generalized Schrödinger bridge (NCGSB). This new approach overcomes limitations of previous methods by allo…
-
Lyapunov-based energy matching offers new perspective on generative models
Researchers have introduced a novel framework for generative models that utilizes a single, time-independent energy function to drive sample generation. This approach unifies training and sampling phases by framing them…
-
New methods enhance robust optimization with ensemble models and worst-case distribution analysis
Researchers have developed new methods for distributionally robust optimization, a technique that accounts for uncertainty in data distributions. One approach, Ensemble Distributionally Robust Bayesian Optimization, use…
-
Quantitative Laplace-type convergence results for exponential probability measures studied
This paper explores quantitative Laplace-type convergence results for exponential probability measures, focusing on norm-like potentials. It establishes bounds between measures $\pi_\varepsilon$ and $\pi_0$ using Wasser…
-
AI models enable whole-cell segmentation in histology images
Researchers have developed two novel AI approaches for histopathology image analysis. One method, VitaminP, uses cross-modal learning to enable whole-cell segmentation from standard H&E stained images by transferring in…
-
Geometric tempering for gradient flow dynamics explored in new arXiv paper
Researchers have investigated geometric tempering as a method for sampling from probability distributions, framing it as an optimization problem. Their work analyzes the impact of using a sequence of moving targets on W…