The field of time series forecasting and digital twins is rapidly advancing, with a focus on developing innovative methods for predicting complex patterns and behaviors in various domains. Recent research has explored the use of deep learning techniques, such as recurrent neural networks and transformers, to improve forecasting accuracy and efficiency. Additionally, there is a growing interest in developing digital twin technologies that can simulate and predict the behavior of complex systems, such as manufacturing processes and materials. Noteworthy papers in this area include the Adaptive Digital Twin of Sheet Metal Forming, which presents a novel framework for adaptive digital twins in nonlinear manufacturing systems, and the Epistemic Error Decomposition for Multi-step Time Series Forecasting, which rethinks the traditional bias-variance tradeoff in recursive and direct forecasting strategies. Other notable papers, such as Predicting Grain Growth in Polycrystalline Materials Using Deep Learning Time Series Models and R-Tuning: Wavelet-Decomposed Replay and Semantic Alignment for Continual Adaptation of Pretrained Time-Series Models, demonstrate the potential of deep learning and digital twin technologies in advancing our understanding and prediction of complex phenomena.
Advancements in Time Series Forecasting and Digital Twins
Sources
Adaptive Digital Twin of Sheet Metal Forming via Proper Orthogonal Decomposition-Based Koopman Operator with Model Predictive Control
Epistemic Error Decomposition for Multi-step Time Series Forecasting: Rethinking Bias-Variance in Recursive and Direct Strategies
Predicting Grain Growth in Polycrystalline Materials Using Deep Learning Time Series Models
Small Vocabularies, Big Gains: Pretraining and Tokenization in Time Series Models
R-Tuning: Wavelet-Decomposed Replay and Semantic Alignment for Continual Adaptation of Pretrained Time-Series Models
Moirai 2.0: When Less Is More for Time Series Forecasting
ReCast: Reliability-aware Codebook Assisted Lightweight Time Series Forecasting
Conformal Online Learning of Deep Koopman Linear Embeddings
Optimal Look-back Horizon for Time Series Forecasting in Federated Learning
APT: Affine Prototype-Timestamp For Time Series Forecasting Under Distribution Shift
Adapformer: Adaptive Channel Management for Multivariate Time Series Forecasting
Multi-layer Stack Ensembles for Time Series Forecasting
Attention-Based Feature Online Conformal Prediction for Time Series
Enhancing Forex Forecasting Accuracy: The Impact of Hybrid Variable Sets in Cognitive Algorithmic Trading Systems