Non-Stationary Time Series

Follow this topic Ask about this topic

Non-stationary time series have changing properties such as their mean, variance, seasonality, or relationships between variables, which can make forecasts and causal conclusions unreliable. Common approaches include change-point segmentation and regime-switching models, time-varying state-space or Markov models, decomposition into trends and residuals, and adaptive neural methods. These techniques support forecasting, anomaly detection, causal discovery, and signal restoration, and are evaluated on settings with structural breaks, drift, changing volatility, and transient events. A central tension is how to adapt to genuine changes without overfitting noise or erasing useful signal structure; benchmarks also need strong classical baselines and diverse dynamics.

Latest papers

24 papers5 papers in the last four weeks, against 2 the four weeks before. 0.0% of all new papers.

Time-partitioning Gaussian Process Mixture Model
Oct 14, 2026Guilherme Piantino, Hedibert F. LopesTime Series Modeling Non-Stationary Time Series
Insper Institute of Education and Research, São Paulo
MORA: Modeling Observed Changes for Drift-Robust Time-Series Anomaly Detection
Oct 7, 2026Xudong Mou, Tiejun Wang, Rui Wang +5Time-Series Anomaly Detection Non-Stationary Time Series
Beihang University, Beijing, China · Beihang Hangzhou Innovation Institute, Hangzhou, China · China University of Geosciences (Beijing), Beijing, China
StaFIR: Convex Learning of Stationarity-Aware Causal Filters
Oct 5, 2026Lorena Egger, Mathis LingerTime Series Forecasting Financial Time Series
Drakai Capital Paris
In a Streaming World, Should You Stand Still? A Comprehensive Benchmark of Anomaly Detection in Streams
Sep 30, 2026Magali Parrino, Antoine Ajenjo, Emmanuel Remy +3Online Anomaly Detection Concept Drift
EDF R&D Chatou, France · ENS, PSL University, CNRS, Inria Paris, France · DI ENS, ENS, PSL University, CNRS, Inria Paris, France
Cyclostationary Phase Conditioning for Medical Time Series Diffusion
Sep 28, 2026Samuel Ruiperez-Campillo, Michele Copetti, Jorge da Silva Goncalves +3Time Series Imputation Diffusion-Based Inverse Problems
Dept. of Computer Science ETH Zurich
DrafTS: Time-Aware Decomposition with Residual Correction for Time Series Modeling
Sep 27, 2026Yiqiu Liu, Siru Zhong, Zhiguang Wang +2Time Series Decomposition Non-Stationary Time Series
The Hong Kong University of Science and Technology (Guangzhou), China · Abel AI Lab, USA · Squirrel Ai Learning, USA
Beyond Stationarity in Time Series: Discovering Causal Structures and Latent Regimes via Markov Blankets
Sep 4, 2026Lei Zan, Charles K. Assaad, Emilie Devijver +1Time-Series Causal Discovery Non-Stationary Time Series
Univ. Grenoble Alpes, CNRS, Grenoble INP, LIG, Grenoble, France · Sorbonne Université, INSERM, Institut Pierre Louis d’Epidémiologie et de Santé Publique, F75012, Paris, France
End-to-end Early Classification of Time Series in Non-Stationary Environments
Aug 20, 2026Aurélien Renault, Alexis Bondu, Antoine Cornuéjols +1Time Series Classification Non-Stationary Time Series
Orange Research & AgroParisTech Paris, France · Orange Research Paris, France · AgroParisTech UMR MIA-Paris Palaiseau, France
Identifiable Markov Switching Models with Instantaneous Effects and Exponential Families
Jun 1, 2026Roel Hulsman, Carles Balsells-Rodas, Sara MagliacaneTime-Series Causal Discovery Regime-Switching Dynamical Systems
University of Amsterdam · 2Adyen · 3Imperial College London +1
MedMamba: Multi-View State Space Models with Adaptive Graph Learning for Medical Time Series Classification
May 24, 2026Da Zhang, Bingyu Li, Zhiyuan Zhao +3Graph Structure Learning Selective SSMs
Northwest Polytechnical University · TeleAI, China Telecom · TeleAI +1
Time Segmented Beamforming via Dynamic Programming: Theory and Implementation
May 24, 2026Manan Mittal, Ryan M. Corey, Diego Cuji +2Dynamic Programming Non-Stationary Time Series
Department of Electrical and Computer Engineering, Stony Brook University, Stony Brook, NY 11794 USA · Department of Electrical and Computer Engineering, University of Illinois, Chicago, IL 60607 USA · Department of Electrical and Computer Engineering, University of Massachusetts Dartmouth, Dartmouth, MA 02747 USA +1
Learning Time-Inhomogeneous Markov Dynamics in Financial Time Series via Neural Parameterization
May 6, 2026Jan Rovirosa, Jesse SchmolzeMarkov Models Financial Time Series
University of Wisconsin – Madison, Madison, WI, USA
Dynamic Vine Copulas: Detecting and Quantifying Time-Varying Higher-Order Interactions
May 4, 2026Houman Safaai, Alessandro Marin VargasCopula Models Non-Stationary Time Series
Kempner Institute at Harvard University · Stanford University
PAMod: Modeling Cyclical Shifts via Phase-Amplitude Modulation for Non-stationary Time Series Forecasting
May 1, 2026Yingbo Zhou, Yutong Ye, Shuhao Li +5Distribution Shift Time Series Forecasting
Fudan University · Beihang University · Beijing University of Posts and Telecommunications
TTCD:Transformer Integrated Temporal Causal Discovery from Non-Stationary Time Series Data
Apr 27, 2026Omar Faruque, Sahara Ali, Xue Zheng +1Time-Series Causal Discovery Causal Structure Learning
Department of Information Systems University of Maryland, Baltimore County Baltimore, Maryland, USA · Department of Data Science University of North Texas Denton, Texas, USA · Climate Science Section Lawrence Livermore National Laboratory Livermore, California, USA
Non-Stationarity in the Embedding Space of Time Series Foundation Models
Apr 6, 2026Jinmyeong Choi, Brad Shook, Artur DubrawskiTime-Series Foundation Models Non-Stationary Time Series
Auton Lab Robotics Institute Carnegie Mellon University Pittsburgh, PA 15213, USA
Improving Model Performance by Adapting the KGE Metric to Account for System Non-Stationarity
Apr 5, 2026M Jawad, HV Gupta, YH Wang +3Hydrology Non-Stationary Time Series
Department of Hydrology and Atmospheric Sciences, The University of Arizona, Tucson, AZ, USA, 85721. · Earth and Environmental Science Area, Lawrence Berkeley National Lab, Berkeley, CA
Seeking SOTA: Time-Series Forecasting Must Adopt Taxonomy-Specific Evaluation to Dispel Illusory Gains
Mar 16, 2026Raeid Saqur, Christoph Bergmeir, Blanka Horvath +3Time Series Forecasting Time Series Modeling
Dept. of Computer Science, University of Toronto · Dept. of Mathematics, University of Oxford · Vector Institute +5
The Temporal Markov Transition Field
Mar 9, 2026Michael LeznikTime Series Representation Learning Non-Stationary Time Series
Aristocrat Leisure Limited, London
WaveSSM: Multiscale State-Space Models for Non-stationary Signal Attention
Feb 25, 2026Ruben Solozabal, Velibor Bojkovic, Hilal Alquabeh +3State Space Models Multiscale Time-Series Modeling
MBZUAI · RIKEN AIP
The Stationarity Bias: Stratified Stress-Testing for Time-Series Imputation in Regulated Dynamical Systems
Feb 17, 2026Amirreza Dolatpour Fathkouhi, Alireza Namazi, Heman ShakeriContinuous Glucose Monitoring Time Series Imputation
Department of Computer Science Center for Diabetes Technology University of Virginia Charlottesville, Virginia, USA · School of Data Science Center for Diabetes Technology University of Virginia Charlottesville, Virginia, USA
DCD: Decomposition-based Causal Discovery from Autocorrelated and Non-Stationary Temporal Data
Feb 1, 2026Muhammad Hasan Ferdous, Ian Stockwell, Md Osman GaniTime-Series Causal Discovery Causal Discovery
Causal AI Lab, Department of Information Systems, University of Maryland, Baltimore County, Baltimore, MD, USA
Cyclic MPDR Beamforming for Suppression of Almost-Cyclostationary Acoustic Interference
Oct 21, 2025Giovanni Bologni, Martin Bo M\oller, Richard Heusdens +1Speech Enhancement Non-Stationary Time Series
Signal Processing Systems Group, Delft University of Technology, Delft, The Netherlands · Bang & Olufsen, Struer, 7600, Denmark · Netherlands Defence Academy, 1781 AC Den Helder, The Netherlands
Robust Streaming PCA
Feb 8, 2019Daniel Bienstock, Minchan Jeong, Apurv Shukla +1Principal Component Analysis Streaming Algorithms
Graduate School of AI KAIST

Oct 14, 2026J/K move · Enter open · S save

Time-partitioning Gaussian Process Mixture Model

Guilherme Piantino, Hedibert F. Lopes

Insper Institute of Education and Research, São Paulo

Time Series Modeling Non-Stationary Time Series Change-Point Detection Gaussian Processes Bayesian Inference

Open paper AskPDF arXiv

Nonstationarity and heterogeneous dynamics are common in financial and macroeconomic time series. Volatility alternates between calm and turbulent regimes, interest rates follow monetary-policy cycles, and major events trigger persistent structural shifts. We propose the Time-partitioning Gaussian Process Mixture Model (TPGPMM), a Bayesian nonparametric framework for regime identification and nonstationary prediction that infers such regimes and their within-regime dependence from data, without Markovian assumptions. It combines (i) an adaptive Pitman-Yor process prior on ordered compositions, mapped to block lengths with contiguity enforced by construction, and (ii) regime-specific Gaussian processes for temporal dependence, yielding a nonstationary predictive distribution. By inferring its discount parameter, the prior adapts to homogeneous and highly heterogeneous regime lengths. Unlike sequential or Markov-switching methods, it flags persistent structural shifts as change points rather than reacting to transient shocks. A split-merge Metropolis-within-Gibbs sampler with exact Gibbs updates of regime boundaries jointly learns change points and hyperparameters in batch, quantifying uncertainty over the partition and within-regime dynamics. With an RBF kernel and a fixed, data-informed noise level, TPGPMM produces a sparse, confident segmentation. On simulated data it recovers the true regimes, and on the CBOE Volatility Index and the 10-year US Treasury yield it identifies regimes aligned with documented market and monetary-policy events. For prediction, variants estimating the noise level fit better than change-point baselines and fit and forecast better than the main specification, at the cost of a less sparse or less event-aligned segmentation; with a regime mean, they fit better than a fitted stationary Gaussian process and have lower forecast RMSE except at the shortest horizons.