Defines outcome-conformant synthesis as exact closed-form generation of relational data matching declared aggregates via Gamma conditional-sum sampling, introduces SpecBench for measuring conformance, and shows it is orthogonal to fidelity.
hub
Real-valued (Medical) Time Series Generation with Recurrent Conditional GANs
21 Pith papers cite this work. Polarity classification is still indexing.
21
Pith papers citing it
abstract
Generative Adversarial Networks (GANs) have shown remarkable success as a framework for training models to produce realistic-looking data. In this work, we propose a Recurrent GAN (RGAN) and Recurrent Conditional GAN (RCGAN) to produce realistic real-valued multi-dimensional time series, with an emphasis on their application to medical data. RGANs make use of recurrent neural networks in the generator and the discriminator. In the case of RCGANs, both of these RNNs are conditioned on auxiliary information. We demonstrate our models in a set of toy datasets, where we show visually and quantitatively (using sample likelihood and maximum mean discrepancy) that they can successfully generate realistic time-series. We also describe novel evaluation methods for GANs, where we generate a synthetic labelled training dataset, and evaluate on a real test set the performance of a model trained on the synthetic data, and vice-versa. We illustrate with these metrics that RCGANs can generate time-series data useful for supervised training, with only minor degradation in performance on real test data. This is demonstrated on digit classification from 'serialised' MNIST and by training an early warning system on a medical dataset of 17,000 patients from an intensive care unit. We further discuss and analyse the privacy concerns that may arise when using RCGANs to generate realistic synthetic medical time series data.
hub tools
citation-role summary
background 1
citation-polarity summary
verdicts
UNVERDICTED 21roles
background 1polarities
background 1representative citing papers
CaTSG is a unified diffusion model for causal time series generation that handles observational, interventional, and counterfactual tasks via backdoor adjustment and abduction-action-prediction.
EMGFlow is the first application of flow matching to synthesize sEMG data, outperforming GAN and diffusion baselines in fidelity, distributional metrics, and downstream gesture recognition utility under TSTR evaluation.
Introduces SOCK (SOft Competing Kernels), a differentiable random convolutional feature map, to train generative models of financial time series via feature matching and shows outperformance over signature and diffusion baselines on small-sample datasets.
ReGeN decomposes references into periodic, stochastic, and causal components to generate synthetic multivariate time series that preserve domain structure and support improved forecasting in low-data settings.
FedEHR-Gen is a federated two-stage autoencoder plus TCVAE system that aligns latent spaces via layer-wise matching and uses distribution-aware aggregation to produce synthetic EHR time-series data matching centralized performance on eICU and MIMIC-III.
PrismFlow augments flow matching with residual dynamical experts and a winner-take-all objective to reduce spectral distortion and improve mode coverage in time-series generation.
GenTS is a modular benchmark library providing unified data pipelines, generative models, and evaluation metrics for time series synthesis, forecasting, and imputation, with open-source code and initial benchmarking experiments.
TabSCM produces causally consistent tabular data by orienting a CPDAG into a DAG, fitting root marginals with KDE, and using conditional diffusion plus trees for child nodes, outperforming GANs and diffusion baselines on fidelity, utility, and privacy across seven datasets.
A new framework evaluates utility of synthetic mobility trajectories while a membership inference attack reveals privacy vulnerabilities in generative models thought to be safe.
Fed-CausalDiff proposes decoupled synchronization in a federated causal diffusion model to improve do-simulation and policy-value estimation across heterogeneous decentralized datasets.
UPLOTS proposes a unified prompt-guided pretrained transformer for generating constrained time-series data across diverse domains using dynamic multi-dataset loss re-weighting.
Introduces a sequential forward-backward diffusion framework that generates adapted time series by conditioning on prior history, with a parallelizable score-matching objective and statistical guarantees for ReLU networks.
AugMask is a plug-and-play training framework that lets diffusion models on incomplete tabular data use stochastic augmentation for conditioning and observed-only supervision, outperforming missing-aware baselines via a Rao-Blackwellized objective.
MOSAIC structures LLM-based model selection via memory-grounded blueprints and failure-aware RL, reporting gains in performance and traceability on financial time-series tasks over AutoML and agent baselines.
Hybrid CoMeTS-GAN plus diffusion model generates multivariate financial time series claimed to better reproduce stylized facts and inter-asset correlations than prior generative methods.
MSDformer introduces a multi-scale discrete transformer that tokenizes time series at multiple scales and models them autoregressively in discrete space, claiming superior performance over prior DTM methods with rate-distortion theoretical support.
Off-the-shelf image diffusion models can be repurposed to create synthetic structured data capable of inducing ground truth drift in machine pipelines.
An MCMC framework enforces empirical transition laws on GAN outputs to reduce temporal drift in synthetic multivariate time series.
TriHead-GAN is a GAN framework whose triple-head discriminator supervises distributional authenticity, cross-variable dependency via regression, and temporal smoothness via adjacent-difference prediction for carbon emission time series.
CGANs with LSTM generator can produce synthetic crypto price series that reproduce temporal patterns and preserve market trends and dynamics.
citing papers explorer
-
Declarative Outcome-Conformant Synthesis: Exact, Closed-Form Specification Satisfaction and a Conformance Benchmark
Defines outcome-conformant synthesis as exact closed-form generation of relational data matching declared aggregates via Gamma conditional-sum sampling, introduces SpecBench for measuring conformance, and shows it is orthogonal to fidelity.
-
Causal Time Series Generation via Diffusion Models
CaTSG is a unified diffusion model for causal time series generation that handles observational, interventional, and counterfactual tasks via backdoor adjustment and abduction-action-prediction.
-
EMGFlow: Robust and Efficient Surface Electromyography Synthesis via Flow Matching
EMGFlow is the first application of flow matching to synthesize sEMG data, outperforming GAN and diffusion baselines in fidelity, distributional metrics, and downstream gesture recognition utility under TSTR evaluation.
-
Generating Financial Time Series by Matching Random Convolutional Features
Introduces SOCK (SOft Competing Kernels), a differentiable random convolutional feature map, to train generative models of financial time series via feature matching and shows outperformance over signature and diffusion baselines on small-sample datasets.
-
REGEN: Reference-Guided Synthetic Multivariate Time Series Generation for Forecasting
ReGeN decomposes references into periodic, stochastic, and causal components to generate synthetic multivariate time series that preserve domain structure and support improved forecasting in low-data settings.
-
FedEHR-Gen: Federated Synthetic Time-Series EHR Generation via Latent Space Alignment and Distribution-Aware Aggregation
FedEHR-Gen is a federated two-stage autoencoder plus TCVAE system that aligns latent spaces via layer-wise matching and uses distribution-aware aggregation to produce synthetic EHR time-series data matching centralized performance on eICU and MIMIC-III.
-
PrismFlow: Residual Dynamics for Flow Matching in Time-Series Generation
PrismFlow augments flow matching with residual dynamical experts and a winner-take-all objective to reduce spectral distortion and improve mode coverage in time-series generation.
-
GenTS: A Comprehensive Benchmark Library for Generative Time Series Models
GenTS is a modular benchmark library providing unified data pipelines, generative models, and evaluation metrics for time series synthesis, forecasting, and imputation, with open-source code and initial benchmarking experiments.
-
TabSCM: A practical Framework for Generating Realistic Tabular Data
TabSCM produces causally consistent tabular data by orienting a CPDAG into a DAG, fitting root marginals with KDE, and using conditional diffusion plus trees for child nodes, outperforming GANs and diffusion baselines on fidelity, utility, and privacy across seven datasets.
-
A Dual Perspective on Synthetic Trajectory Generators: Utility Framework and Privacy Vulnerabilities
A new framework evaluates utility of synthetic mobility trajectories while a membership inference attack reveals privacy vulnerabilities in generative models thought to be safe.
-
Fed-CausalDiff: Decoupled Synchronization for Federated Do-Simulation and Policy Evaluation
Fed-CausalDiff proposes decoupled synchronization in a federated causal diffusion model to improve do-simulation and policy-value estimation across heterogeneous decentralized datasets.
-
UPLOTS: A Unified Pretrained Language Model for Constrained Time-series Generation
UPLOTS proposes a unified prompt-guided pretrained transformer for generating constrained time-series data across diverse domains using dynamic multi-dataset loss re-weighting.
-
Diffusion Models for Adaptive Sequential Data Generation
Introduces a sequential forward-backward diffusion framework that generates adapted time series by conditioning on prior history, with a parallelizable score-matching objective and statistical guarantees for ReLU networks.
-
AugMask: Training Diffusion Models on Incomplete Tabular Data via Stochastic Augmentation and Masking
AugMask is a plug-and-play training framework that lets diffusion models on incomplete tabular data use stochastic augmentation for conditioning and observed-only supervision, outperforming missing-aware baselines via a Rao-Blackwellized objective.
-
MOSAIC: Modular Orchestration for Structured Agentic Intelligence and Composition
MOSAIC structures LLM-based model selection via memory-grounded blueprints and failure-aware RL, reporting gains in performance and traceability on financial time-series tasks over AutoML and agent baselines.
-
High-Quality Synthetic Financial Time-Series using a GAN-Diffusion Framework
Hybrid CoMeTS-GAN plus diffusion model generates multivariate financial time series claimed to better reproduce stylized facts and inter-asset correlations than prior generative methods.
-
MSDformer: Multi-scale Discrete Transformer For Time Series Generation
MSDformer introduces a multi-scale discrete transformer that tokenizes time series at multiple scales and models them autoregressively in discrete space, claiming superior performance over prior DTM methods with rate-distortion theoretical support.
-
Repurposing Image Diffusion Models for Adversarial Synthetic Structured Data: A Case Study of Ground Truth Drift
Off-the-shelf image diffusion models can be repurposed to create synthetic structured data capable of inducing ground truth drift in machine pipelines.
-
Preserving Temporal Dynamics in Time Series Generation
An MCMC framework enforces empirical transition laws on GAN outputs to reduce temporal drift in synthetic multivariate time series.
-
TriHead-GAN: A Generative Adversarial Network with Triple-Head Discriminator for Carbon Emission Time Series Generation
TriHead-GAN is a GAN framework whose triple-head discriminator supervises distributional authenticity, cross-variable dependency via regression, and temporal smoothness via adjacent-difference prediction for carbon emission time series.
-
Synthetic data in cryptocurrencies using generative models
CGANs with LSTM generator can produce synthetic crypto price series that reproduce temporal patterns and preserve market trends and dynamics.