Non-autoregressive Conditional Diffusion Models for Time Series Prediction
원문 초록
Recently, denoising diffusion models have led to significant breakthroughs in the generation of images, audio and text. However, it is still an open question on how to adapt their strong modeling ability to model time series. In this paper, we propose TimeDiff, a non-autoregressive diffusion model that achieves high-quality time series prediction with the introduction of two novel conditioning mechanisms: future mixup and autoregressive initialization. Similar to teacher forcing, future mixup allows parts of the ground-truth future predictions for conditioning, while autoregressive initialization helps better initialize the model with basic time series patterns such as short-term trends. Extensive experiments are performed on nine real-world datasets. Results show that TimeDiff consistently outperforms existing time series diffusion models, and also achieves the best overall performance across a variety of the existing strong baselines (including transformers and FiLM).
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
Cross-Domain Conditional Diffusion Models for Time Series Imputation
Predicting Graph Structure via Adapted Flux Balance Analysis
SEED: A Structural Encoder for Embedding-Driven Decoding in Time Series Prediction with LLMs
Time-series surrogates from energy consumers generated by machine learning approaches for long-term forecasting scenarios
MS-TVNet:A Long-Term Time Series Prediction Method Based on Multi-Scale Dynamic Convolution
Neural MJD: Neural Non-Stationary Merton Jump Diffusion for Time Series Prediction
토론
이 논문과 연결된 포럼 글이 아직 없습니다.