ScoreGrad: Multivariate Probabilistic Time Series Forecasting with Continuous Energy-based Generative Models

Yan, Tijin; Zhang, Hongwei; Zhou, Tong; Zhan, Yufeng; Xia, Yuanqing

doi:10.48550/arxiv.2106.10121

Cited by 8 publications

(7 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To yield predictive distribution for multivariate TSF, TLAE [69] implements nonlinear transformation by replacing matrix factorization with encoder-decoder architecture and temporal deep temporal latent model. Another line of generative methods for TSF focus on energy-based models (EBMs), such as TimeGrad [74] and ScoreGrad [105]. EBMs do not restrict the tractability of the normalizing constants [105].…”

Section: A1 Time Series Forecastingmentioning

confidence: 99%

See 1 more Smart Citation

Generative Time Series Forecasting with Diffusion, Denoise, and Disentanglement

Li¹,

Lu²,

Wang³

et al. 2023

Preprint

View full text Add to dashboard Cite

Time series forecasting has been a widely explored task of great importance in many applications. However, it is common that real-world time series data are recorded in a short time period, which results in a big gap between the deep model and the limited and noisy time series. In this work, we propose to address the time series forecasting problem with generative modeling and propose a bidirectional variational auto-encoder (BVAE) equipped with diffusion, denoise, and disentanglement, namely D 3 VAE. Specifically, a coupled diffusion probabilistic model is proposed to augment the time series data without increasing the aleatoric uncertainty and implement a more tractable inference process with BVAE. To ensure the generated series move toward the true target, we further propose to adapt and integrate the multiscale denoising score matching into the diffusion process for time series forecasting. In addition, to enhance the interpretability and stability of the prediction, we treat the latent variable in a multivariate manner and disentangle them on top of minimizing total correlation. Extensive experiments on synthetic and real-world data show that D 3 VAE outperforms competitive algorithms with remarkable margins. Our implementation is available at https://github.com/ PaddlePaddle/PaddleSpatial/tree/main/research/D3VAE.

show abstract

Section: A1 Time Series Forecastingmentioning

confidence: 99%

“…Another line of generative methods for TSF focus on energy-based models (EBMs), such as TimeGrad [74] and ScoreGrad [105]. EBMs do not restrict the tractability of the normalizing constants [105]. Though flexible, the unknown normalizing constant makes the training of EBMs particularly difficult.…”

Section: A1 Time Series Forecastingmentioning

confidence: 99%

Generative Time Series Forecasting with Diffusion, Denoise, and Disentanglement

Li¹,

Lu²,

Wang³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…In Ref. [117], a general framework based on continuous energy-based generative models for time series forecasting is established. The training process at each step is composed of a time series feature extraction module and a conditional SDE based score matching module.…”

Section: Appendix D: Solving Reverse-time Stochastic Differential Equ...mentioning

confidence: 99%

Quantum Quantile Mechanics: Solving Stochastic Differential Equations for Generating Time‐Series

Paine

Elfving²,

Kyriienko

2023

Adv Quantum Tech

View full text Add to dashboard Cite

A quantum algorithm is proposed for sampling from a solution of stochastic differential equations (SDEs). Using differentiable quantum circuits (DQCs) with a feature map encoding of latent variables, the quantile function is represented for an underlying probability distribution and samples extracted as DQC expectation values. Using quantile mechanics the system is propagated in time, thereby allowing for time‐series generation. The method is tested by simulating the Ornstein‐Uhlenbeck process and sampling at times different from the initial point, as required in financial analysis and dataset augmentation. Additionally, continuous quantum generative adversarial networks (qGANs) are analyzed, and the authors show that they represent quantile functions with a modified (reordered) shape that impedes their efficient time‐propagation. The results shed light on the connection between quantum quantile mechanics (QQM) and qGANs for SDE‐based distributions, and point the importance of differential constraints for model training, analogously with the recent success of physics informed neural networks.

show abstract

“…In Ref. [103], a general framework based on continuous energybased generative models for time series forecasting is established. The training process at each step is composed of a time series feature extraction module and a conditional SDE based score matching module.…”

Section: Boundary Handling For Pdesmentioning

confidence: 99%

Quantum Quantile Mechanics: Solving Stochastic Differential Equations for Generating Time-Series

Paine,

Elfving,

Kyriienko

2021

Preprint

View full text Add to dashboard Cite

We propose a quantum algorithm for sampling from a solution of stochastic differential equations (SDEs). Using differentiable quantum circuits (DQCs) with a feature map encoding of latent variables, we represent the quantile function for an underlying probability distribution and extract samples as DQC expectation values. Using quantile mechanics we propagate the system in time, thereby allowing for time-series generation. We test the method by simulating the Ornstein-Uhlenbeck process and sampling at times different from the initial point, as required in financial analysis and dataset augmentation. Additionally, we analyse continuous quantum generative adversarial networks (qGANs), and show that they represent quantile functions with a modified (reordered) shape that impedes their efficient time-propagation. Our results shed light on the connection between quantum quantile mechanics (QQM) and qGANs for SDE-based distributions, and point the importance of differential constraints for model training, analogously with the recent success of physics informed neural networks.

show abstract

ScoreGrad: Multivariate Probabilistic Time Series Forecasting with Continuous Energy-based Generative Models

Cited by 8 publications

References 45 publications

Generative Time Series Forecasting with Diffusion, Denoise, and Disentanglement

Generative Time Series Forecasting with Diffusion, Denoise, and Disentanglement

Quantum Quantile Mechanics: Solving Stochastic Differential Equations for Generating Time‐Series

Quantum Quantile Mechanics: Solving Stochastic Differential Equations for Generating Time-Series

Contact Info

Product

Resources

About