Results 51 to 60 of about 3,027 (198)

𝐿₁-distortion of Wasserstein metrics: A tale of two dimensions

Transactions of the American Mathematical Society, Series B, 2023
By discretizing an argument of Kislyakov, Naor and Schechtman proved that the 1-Wasserstein metric over the planar grid { 0 ,
Baudier, F., Gartland, C., Schlumprecht, Th.   +2 more
openaire   +2 more sources

Hybrid physics–data‐driven modeling for sea ice thermodynamics and transfer learning

Quarterly Journal of the Royal Meteorological Society, EarlyView.
Icepack–NN, a machine‐learning‐based hybrid version of the sea‐ice column model Icepack, is developed to correct state‐dependent forecast errors arising from misspecified snow thermodynamics, using neural networks applied online within the physical model.
G. De Cillis, A. Carrassi, J. Brajard, L. Bertino, M. Broccoli, D. Iovino, T.S. Finn, M. Bocquet   +7 more
wiley   +1 more source

Partial Gromov-Wasserstein Metric

Published at ICLR ...
Yikun Bai, Rocio Diaz Martin, Abihith Kothapalli, Hengrong Du, Xinran Liu, Soheil Kolouri   +5 more
openaire   +3 more sources

Machine learning‐driven advances in carbon‐based quantum dots: Opportunities accompanied by challenges

Responsive Materials, EarlyView.
Machine learning provides a unifying framework to connect structure, fluorescence properties, and applications of carbon‐based quantum dots. This review highlights how data‐driven strategies enable fluorescence regulation, reveal underlying mechanisms, and accelerate the rational design of functional carbon dots.
Liangfeng Chen, Siwei Yang, Xiangqian Cui, Ziheng Liu, Hang Wang, Yongqiang Li, Peng He, Miao Wang, Guqiao Ding   +8 more
wiley   +1 more source

Residual Adversarial Subdomain Adaptation Network Based on Wasserstein Metrics for Intelligent Fault Diagnosis of Bearings

Applied Sciences
Subdomain adaptation plays a significant role in the field of bearing fault diagnosis. It effectively aligns the pertinent distributions across subdomains and addresses the frequent issue of lacking local category information in domain adaptation ...
Haichao Cai, Bo Yang, Yujun Xue, Yanwei Xu   +3 more
doaj   +1 more source

Detecting Plateau Zokor (Eospalax baileyi) Mounds in UAV Imagery of Alpine Meadows Using Deep Learning Algorithms

Remote Sensing in Ecology and Conservation, EarlyView.
We developed PZM‐YOLO to automatically detect plateau zokor mounds in UAV imagery of alpine meadows. The model achieved reliable detection of small and densely distributed mounds under complex backgrounds, outperforming the baseline YOLOv5s. This framework supports mound counting, mound position, rodent impact assessment, and grassland restoration ...
Yang Yang, Limin Hua, Lianguo Wang, Jiazhen Li, Rui Dong, Siwei Yang   +5 more
wiley   +1 more source

Dataset Distillation via the Wasserstein Metric

2025 IEEE/CVF International Conference on Computer Vision (ICCV)
Dataset Distillation (DD) aims to generate a compact synthetic dataset that enables models to achieve performance comparable to training on the full large dataset, significantly reducing computational costs. Drawing from optimal transport theory, we introduce WMDD (Wasserstein Metric-based Dataset Distillation), a straightforward yet powerful method ...
Haoyang Liu 0001, Yijiang Li, Tiancheng Xing, Peiran Wang, Vibhu Dalal, Luwei Li, Jingrui He, Haohan Wang   +7 more
openaire   +2 more sources

Learning Across the Divide: Understanding Knowledge Sharing Through Petrographic Analysis on Ceramics From the Rhine‐Meuse Delta During the Middle to Late Neolithic Transition (3400–2200 bce)

Archaeometry, EarlyView.
ABSTRACT Vlaardingen (VL) communities on the Dutch West coast (3400–2200 bce) are part of a unique, long‐term continuity in the European Neolithic. Despite large‐scale changes in European populations during the Neolithic, the genomic diversity and cultural practices of VL communities can be retraced to the Mesolithic.
Jisca de Bruin, Erik Kroon, Lasse van den Dikkenberg, Dennis Braekmans   +3 more
wiley   +1 more source

Exploring the Limitations of Federated Learning: A Novel Wasserstein Metric-Based Poisoning Attack on Traffic Sign Classification

IEEE Access
Federated Learning (FL) enhances privacy but remains vulnerable to model poisoning attacks, where an adversary manipulates client models to upload poisoned updates during training, thereby compromising the overall FL model.
Suzan Almutairi, Ahmed Barnawi
doaj   +1 more source

Geometric standardized mean difference and its application to meta‐analysis

British Journal of Mathematical and Statistical Psychology, EarlyView.
Abstract The standardized mean difference (SMD) is a widely used measure of effect size, particularly common in psychology, clinical trials and meta‐analysis involving continuous outcomes. Traditionally, under the equal variance assumption, the SMD is defined as the mean difference divided by a common standard deviation.
Jiandong Shi, Xiaochen Zhang, Lu Lin, Hiu Yee Kwan, Tiejun Tong   +4 more
wiley   +1 more source