On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

169,236 papers · 148 categories

Trend · papers per month

count % of corpus

5 results for “Flow-GAN”

Flow-GAN: Combining Maximum Likelihood and Adversarial Learning in Generative Modelscs.LG

Flow-GAN combines GAN and maximum likelihood for better sample quality and likelihood evaluation.

problem Challenges in evaluating generative models trained by maximum likelihood.

method Proposes Flow-GAN, a generative adversarial network that allows exact likelihood evaluation.

result Hybrid training of Flow-GAN can achieve high held-out likelihoods while maintaining visual fidelity.

Flow-GAN Generative Adversarial Networks Maximum Likelihood Adversarial Learning Likelihood Evaluation

2017-05-24abs ↗pdf ↗

Flows for simultaneous manifold learning and density estimationstat.ML

M-flows learn data manifolds and densities, improving manifold learning and inference.

problem Representing datasets with manifold structure more faithfully.

method Combining normalizing flows, GANs, autoencoders, and energy-based models, with a new training algorithm.

result M-flows learn data manifolds better than standard flows and provide handles for dimensionality reduction.

manifold learning density estimation generative models maximum likelihood normalizing flows

2020-03-31abs ↗pdf ↗

Learning with minibatch Wasserstein : asymptotic and gradient propertiesstat.ML

Paper analyzes minibatch Wasserstein for machine learning applications.

problem Optimal transport distances on large datasets.

method Analysis of minibatch optimal transport.

result Minibatch Wasserstein is equivalent to implicit regularization with desirable properties.

optimal transport machine learning minibatch Wasserstein regularization

2019-10-09abs ↗pdf ↗

Discovering state-parameter mappings in subsurface models using generative adversarial networksphysics.data-an

SPID-GAN learns bidirectional mappings in subsurface models.

problem Challenges in identifying and approximating causal structures in high-dimensional parameter spaces.

method Generative adversarial networks (GANs) for learning cross-domain mappings.

result SPID-GAN achieves satisfactory performance in identifying bidirectional state-parameter mappings.

geophysical modeling subsurface flow GANs state-parameter mappings deep learning

2018-10-30abs ↗pdf ↗

Missing Data Imputation by Reducing Mutual Information with Rectified Flowsstat.ML

Paper proposes a novel method to reduce mutual information for missing data imputation.

problem Missing data imputation in datasets with missingness patterns.

method Iterative minimization of KL divergence between imputed data and missingness mask, using rectified flow training objective.

result The proposed method achieves superior imputation performance on synthetic and real-world datasets.

missing data imputation mutual information rectified flow GAN

2025-05-16abs ↗pdf ↗