Results 51 to 60 of about 24,886 (258)
Mapper Comparison with Wasserstein Metrics
CoRR, 2018 The challenge of describing model drift is an open question in unsupervised learning. It can be difficult to evaluate at what point an unsupervised model has deviated beyond what would be expected from a different sample from the same population. This is particularly true for models without a probabilistic interpretation. One such family of techniques,
openaire +2 more sources
Distribution’s template estimate with Wasserstein metrics
Bernoulli, 2015 In this paper we tackle the problem of comparing distributions of random variables and defining a mean pattern between a sample of random events. Using barycenters of measures in the Wasserstein space, we propose an iterative version as an estimation of the mean distribution.
Boissard, Emmanuel +2 more
openaire +5 more sources
Advanced Science, EarlyView.TarPass provides a rigorous benchmark for target‐aware de novo molecular generation by jointly evaluating protein‐ligand interactions, molecular plausibility, and drug‐likeness on 18 well‐studied targets. Results show that current models often fail to consistently surpass random baseline in target‐specific enrichment, while post hoc multi‐tier virtual ...
Rui Qin +11 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
Advanced Science, EarlyView.Biological rhythms coordinate physiology, from genes to behavior. Study of circadian rhythms in brain tissue is constrained by limited throughput and spatial and temporal information quality. A new platform for high‐throughput, long‐term multiplexed fluorescent live imaging of circadian rhythms in brain slices is introduced.
Marco Ferrari +3 more
wiley +1 more source
Stability of the global attractor under Markov-Wasserstein noise [PDF]
, 2011 We develop a "weak Wa\.zewski principle" for discrete and continuous time dynamical systems on metric spaces having a weaker topology to show that attractors can be continued in a weak sense.
Kell, Martin
core
Advanced Energy Materials, EarlyView.Machine learning interatomic potentials bridge quantum accuracy and computational efficiency for materials discovery. Architectures from Gaussian process regression to equivariant graph neural networks, training strategies including active learning and foundation models, and applications in solid‐state electrolytes, batteries, electrocatalysts ...
In Kee Park +19 more
wiley +1 more source
Optimal Transport for Seismic Full Waveform Inversion
, 2016 Full waveform inversion is a successful procedure for determining properties of the earth from surface measurements in seismology. This inverse problem is solved by a PDE constrained optimization where unknown coefficients in a computed wavefield are ...
Engquist, Bjorn +2 more
core +1 more source
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 , 1 , … , n } 2 \{0,1,\dots , n\}^2 has L 1 L_1 -distortion bounded below by a constant ...
Baudier, F. +2 more
openaire +2 more sources
Advanced Intelligent Discovery, EarlyView.We investigate MACE‐MP‐0 and M3GNet, two general‐purpose machine learning potentials, in materials discovery and find that both generally yield reliable predictions. At the same time, both potentials show a bias towards overstabilizing high energy metastable states. We deduce a metric to quantify when these potentials are safe to use.
Konstantin S. Jakob +2 more
wiley +1 more source