Guest Editors
Prof. Hossein Hassani
Email: hassani@iiasa.ac.at
Affiliation: Cooperation and Transformative Governance Group, Advancing Systems Analysis Program, International Institute for Applied Systems Analysis (IIASA), 2361 Laxenburg, Austria
Homepage:
Research Interests: time series analysis, forecasting, big data, digital twins, signal processing, noise reduction, modelling, statistical analysis
Prof. Emmanuel Silva
Email: emmanuel.silva@gcu.ac.uk
Affiliation: Glasgow School for Business and Society, Glasgow Caledonian University, Glasgow G4 0BA, UK
Homepage:
Research Interests: time series analysis, forecasting, big data, modelling, singular spectrum analysis
Dr. Leila Marvian
Email: leila.marveian@imamreza.ac.ir
Affiliation: Data Lab, Imam Reza University, Mashhad, IRAN 511, Iran
Homepage:
Research Interests: time series analysis, forecasting, ARIMA models, big data, modelling
Summary
Time-series data underpin critical decisions across science, engineering, finance, and public policy, yet modern deployments face nonstationarity, scale, and shifting environments. Rapid progress in probabilistic, causal, and deep sequence modeling now enables richer forecasts and insights, calling for a coherent venue to consolidate advances and best practices.
This Special Issue seeks methodological and applied advances in time-series analysis, modeling, and forecasting. We welcome theory, algorithms, and applications addressing nonstationarity, irregular sampling, missing data, and high dimensionality. Topics include probabilistic forecasting and calibration, representation learning, hybrid physics–ML models, causal inference, and online/streaming decisioning. Submissions must demonstrate rigorous evaluation with statistically principled comparisons (e.g., Diebold–Mariano, Hassani–Silva KS) and report uncertainty, robustness, and reproducibility with open code and data where possible.
· Nonstationarity, regime shifts, change-point detection, and adaptation to distribution shift
· Probabilistic forecasting, calibration, conformal prediction, and decision-theoretic evaluation (CRPS, pinball loss)
· High-dimensional multivariate/graph and spatiotemporal series; state-space and operator-learning methods
· Hybrid physics-informed and mechanistic-ML models; real-time/edge forecasting under resource constraints
· Representation learning for sequences (transformers, SSMs, diffusion) and interpretable/explainable models
· Causal discovery and counterfactual/policy-aware forecasting
· Evaluation, robustness, and reproducibility: loss-based tests, distributional tests, benchmarks, and open code/data
Keywords
time series analysis, modelling, forecasting, noise reduction, signal processing, prediction, filtering, big data
Published Papers
Login
Register
Submit a Paper
Propose a Special lssue