Results 41 to 50 of about 3,027 (198)
MolMiner: Toward Controllable, Three‐Dimensional‐Aware, Fragment‐Based Molecular Design
Advanced Intelligent Discovery, EarlyView.MolMiner is a fragment‐based, geometry‐aware, and order‐agnostic generative model for molecular design with strong inductive biases. Using symmetry‐aware fragment assembly, dynamic three‐dimensional geometry, and multi‐property conditioning, MolMiner enables interpretable and controllable molecular generation.
Raul Ortega‐Ochoa +2 more
wiley +1 more source
(q,p)-Wasserstein GANs: Comparing Ground Metrics for Wasserstein GANs
CoRR, 2019 Generative Adversial Networks (GANs) have made a major impact in computer vision and machine learning as generative models. Wasserstein GANs (WGANs) brought Optimal Transport (OT) theory into GANs, by minimizing the $1$-Wasserstein distance between model and data distributions as their objective function.
Anton Mallasto +3 more
openaire +2 more sources
2D Implementation of Kinetic‐Diffusion Monte Carlo in Eiron
Contributions to Plasma Physics, EarlyView.ABSTRACT Particle‐based kinetic Monte Carlo simulations of neutral particles are one of the major computational bottlenecks in tokamak scrape‐off layer simulations. This computational cost comes from the need to resolve individual collision events in high‐collisional regimes.
Oskar Lappi +3 more
wiley +1 more source
Permutation invariant networks to learn Wasserstein metrics
CoRR, 2020 Understanding the space of probability measures on a metric space equipped with a Wasserstein distance is one of the fundamental questions in mathematical analysis. The Wasserstein metric has received a lot of attention in the machine learning community especially for its principled way of comparing distributions.
Arijit Sehanobish +2 more
openaire +2 more sources
Nonembeddability of persistence diagrams with 𝑝>2 Wasserstein metric [PDF]
Proceedings of the American Mathematical Society, 2021 Persistence diagrams do not admit an inner product structure compatible with any Wasserstein metric. Hence, when applying kernel methods to persistence diagrams, the underlying feature map necessarily causes distortion. We prove that persistence diagrams with the
openaire +3 more sources
Generative Models in Inorganic Crystals Discovery and Inverse Design
ENERGY &ENVIRONMENTAL MATERIALS, EarlyView.Generative inverse‐design samples from the vast inorganic crystal design space by starting from target properties such as band gap, stability, and ion transport. This Review examines the representations, generative models, and validation workflows needed to translate candidate structures into stable, potentially synthesizable materials for applications
Tao Li +5 more
wiley +1 more source
Quantization-based Bounds on the Wasserstein Metric
CoRR23 pages, 8 figures, 7 ...
Jonathan Bobrutsky, Amit Moscovich
openaire +2 more sources
People and Nature, EarlyView.Abstract High aggregate levels of wildlife consumption in cities in Central Africa highlight the need for solutions that balance wildlife protection, local livelihoods and the relational values between people and nature. This study explores the impacts of demand‐ and supply‐side interventions on wild meat consumption through two randomized control ...
Abdoulaye Cisse +2 more
wiley +1 more source
Digital Communications and NetworksThe performance degradation of mega Non-Geostationary Orbit (NGSO) satellite systems caused by co-frequency interference exhibits significant uncertainty, primarily due to the time-varying geometric topology, diverse communication signals, and dynamic ...
Yuanzhi He, Liujing Hu, Li Chen
doaj +1 more source
Canonical Variates in Wasserstein Metric Space
CoRRIn this paper, we address the classification of instances represented by distributions on a vector space rather than single points. We consider classification algorithms based on pairwise distances, specifically, the Wasserstein metric between distributions.
Jia Li, Lin Lin
openaire +2 more sources