On the Evaluation of Generative Adversarial Networks By Discriminative Models

Torfi, Amirsina; Beyki, Mohammadreza; Fox, Edward A.

doi:10.1109/icpr48806.2021.9412214

Cited by 11 publications

(18 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Synthetic data generation can also be utilized to discover new scientific principles by grounding it in biological priors [54]. There have been a good number of models and software developed, such as SynSys, which uses hidden Markov models and regression models initially trained on real datasets to generate synthetic time series data consisting of nested sequences [55]; dahmen2019synsys and corGAN, in which synthetic data is generated by capturing correlations between adjacent medical features in the data representation space [19].…”

Section: Healthcarementioning

confidence: 99%

“…Manuscript submitted to ACM [16] healthcare GAN MLP MIMIC/Sutter (Electronic health record) MMCGAN [17] healthcare & CV GAN CNN chest CT images DeepSynth [18] healthcare & CV GAN CNN rat kidney tissue (microscope image) CorGAN [19] healthcare GAN CNN MIMIC-III dataset, UCI Epileptic Seizure Recognition dataset DAAE [20] healthcare VAE+GAN recurrent autoencoder MIMIC-III, UT Physicians clinical databases HAPNEST [21] healthcare [24] vision GAN deep CGAN MNIST BigGANs [25] vision GAN large scale GAN ImageNet VideoDiff [26] vision diffusion CNN BAIR Robot Pushing, Kinetics-600 VQ-VAE [27] vision VAE PixelCNN ImageNet GIRAFFE [28] vision GAN CNN CompCars, LSUN Churches, and FFHQ Wavegrad [29] TTS diffusion gradient-based sampling LJ Speech TTS-GAN [30] TTS GAN auto-regressive model Tacotron2 Seq-GAN [31] NLP GAN+RL CNN Nottingham dataset BLEURT [32] NLP Language model BERT WebNLG Competition dataset TextGen-RL [33] NLP RL LSTM SynBench [34] NLP conditional Gaussian mixture…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Machine Learning for Synthetic Data Generation: a Review

Lu¹,

Wang²,

Wei³

2023

Preprint

View full text Add to dashboard Cite

Data plays a crucial role in machine learning. However, in real-world applications, there are several problems with data, e.g., data are of low quality; a limited number of data points lead to under-fitting of the machine learning model; it is hard to access the data due to privacy, safety and regulatory concerns. Synthetic data generation offers a promising new avenue, as it can be shared and used in ways that real-world data cannot. This paper systematically reviews the existing works that leverage machine learning models for synthetic data generation. Specifically, we discuss the synthetic data generation works from several perspectives: (i) applications, including computer vision, speech, natural language, healthcare, and business; (ii) machine learning methods, particularly neural network architectures and deep generative models; (iii) privacy and fairness issue. In addition, we identify the challenges and opportunities in this emerging field and suggest future research directions.

show abstract

Section: Healthcarementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Machine Learning for Synthetic Data Generation: a Review

Lu¹,

Wang²,

Wei³

2023

Preprint

View full text Add to dashboard Cite

show abstract

Section: Synthetic Ehr Generationmentioning

confidence: 99%

“…Closely related to this work are recent efforts that leverage deep generative models for synthesizing EHRs [2,8,30]. MedGAN [8] and CorGAN [30] were introduced to generate patient feature matrices. However, these works rely heavily on the performance of a pre-trained autoencoder model to reduce the dimensionality of the latent variable.…”

Section: Synthetic Ehr Generationmentioning

confidence: 99%

“…Thus, generating realistic, but not real data is a key element to advance machine learning for the healthcare community. There have been several distinguished efforts conducted in a variety of domains about synthetic data (EHR) generation [3,4,8,30,35,15,5]. Unfortunately, existing proposed algorithms predominantly adopt a variant of Generative Adversarial Network (GAN) [13,16], auto-encoder [33], or a combination of both.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

MedDiff: Generating Electronic Health Records using Accelerated Denoising Diffusion Model

He¹,

Zhao²,

Xi³

et al. 2023

Preprint

View full text Add to dashboard Cite

Due to patient privacy protection concerns, machine learning research in healthcare has been undeniably slower and limited than in other application domains. High-quality, realistic, synthetic electronic health records (EHRs) can be leveraged to accelerate methodological developments for research purposes while mitigating privacy concerns associated with data sharing. The current state-of-the-art model for synthetic EHR generation is generative adversarial networks, which are notoriously difficult to train and can suffer from mode collapse. Denoising Diffusion Probabilistic Models, a class of generative models inspired by statistical thermodynamics, have recently been shown to generate highquality synthetic samples in certain domains. It is unknown whether these can generalize to generation of large-scale, high-dimensional EHRs. In this paper, we present a novel generative model based on diffusion models that is the first successful application on electronic health records. Our model proposes a mechanism to perform class-conditional sampling to preserve label information. We also introduce a new sampling strategy to accelerate the inference speed. We empirically show that our model outperforms existing stateof-the-art synthetic EHR generation methods.

show abstract