concept drift detection

This study addresses the automatic detection and precise localization of change points in large-scale multivariate time series exhibiting dynamic evolution. We propose a two-stage change point detection method based on feedforward neural networks (FNNs). The method innovatively integrates FNNs into the change point detection framework, incorporating piecewise training, sliding-window error evaluation, and an online recalibration mechanism, along with a dedicated error calibration strategy to ensure estimation consistency under temporal dependence. Theoretically, we establish consistency of the change point localization estimator. Empirically, the method achieves high accuracy in estimating both the number and locations of change points on both synthetic and real-world datasets. Moreover, it supports practical, data-driven selection of hyperparameters, enhancing its applicability in real-world scenarios.

This work addresses the lack of reliable uncertainty quantification in online trend estimation for nonstationary time series by proposing a general online bootstrap method applicable to trend estimators expressed as time-window-weighted sample means—such as exponential smoothing and moving averages. Built upon asymptotic theory, the method provides the first uniform-in-time coverage guarantees for trend inference under nonstationarity, enabling adaptive anomaly detection and A/B testing in streaming data settings. Empirical evaluations demonstrate that the framework achieves well-calibrated uncertainty estimates and scales effectively across diverse nonstationary scenarios, offering a practical and real-time solution for accurate uncertainty quantification in large-scale online time series analysis.

Existing evaluation methods struggle to disentangle whether performance degradation under temporal distribution shifts stems from insufficient model adaptability or increased data difficulty. To address this, this work introduces a novel approach that decouples model adaptability from the inherent difficulty of temporal data for the first time. The authors propose three dynamic metrics based on performance trajectories, which capture the adaptation process through dynamic evaluation and comparative analysis. These new metrics uncover fine-grained adaptation patterns obscured by conventional assessment techniques, substantially enhancing the interpretability and depth of understanding of temporal robustness in machine learning models.

This paper addresses the core challenge in change detection for nonstationary processes—where the post-change distribution evolves over time and is a priori unknown. Method: We propose a robust sequential detection framework based on the least favorable distribution, integrating minimax decision theory, nonstationary statistical modeling, and sequential hypothesis testing to construct an adaptive CUSUM-type algorithm capable of real-time, robust response to dynamic distributional shifts. Contribution/Results: To our knowledge, this is the first work to establish asymptotically minimax-optimal theoretical guarantees for sequential change detection under nonstationarity. Empirical evaluation on real-world public health and military monitoring datasets, as well as extensive simulations, demonstrates that our method reduces average detection delay by 23% compared to conventional approaches, while maintaining stable and controllable false alarm rates—thereby validating both theoretical soundness and practical superiority.

This paper addresses online changepoint detection in nonstationary univariate time series. We propose a Bayesian online method that jointly models time-varying variance and autocorrelation structure within an arbitrary-order autoregressive (AR(p)) framework with time-varying parameters. Our key contributions are: (i) the first integration of time-varying volatility and dynamic autocorrelation into a unified Bayesian changepoint detection framework; and (ii) a scoring-rule-driven recursive parameter update mechanism that preserves memory-aware modeling while enhancing real-time responsiveness. Inference is performed online via the posterior distribution of the current segment length, eliminating the need for fixed sliding windows or offline retraining. Extensive evaluation on real-world datasets across multiple domains demonstrates significant improvements in changepoint localization accuracy and short-term forecasting performance, particularly in capturing complex temporal dependencies and nonstationary evolutionary patterns.

This work addresses the challenge of rapid dynamic shifts in streaming time series caused by abrupt environmental changes or varying input delays. The authors propose a system tensor representation based on Markov parameter sequences, modeling the streaming data as a dynamic mixture of delay systems. By constructing fixed-length tensor summaries that jointly encode system dynamics and input–output delay characteristics, the method enables efficient compression and retrieval of historical patterns through tensor decomposition. Within an online learning framework, the system dynamically selects the optimal submodel to match the current state, achieving strong adaptability to nonstationary time series while maintaining low memory overhead. Experimental results on real-world datasets demonstrate that the proposed approach significantly outperforms existing methods in both prediction accuracy and adaptation speed, particularly under highly nonstationary conditions.

This study addresses the challenge of model performance degradation caused by temporal shifts in feature distributions within real-world time-series data. The authors propose a parameter-free, automated detection method that leverages a regression model to predict sample timestamps and integrates feature importance analysis to identify time-sensitive features. By unifying the quantification of both univariate and multivariate distributional changes across numerical and categorical features, the approach offers strong scalability. Experimental results demonstrate that the method effectively and comprehensively captures a wide range of fundamental drift patterns on both real-world and synthetic datasets, achieving high detection accuracy while maintaining computational efficiency.

This work addresses the challenge of detecting subtle changes in cross-variable and temporal dependency structures within multivariate time series in an online setting. The authors propose a distribution-free, minimally parameterized, and interpretable nonparametric method that recursively estimates Dynamic Mode Decomposition (DMD) operators and monitors their truncated eigenvalue sequences to capture abrupt shifts in dependency patterns. To handle the permutation invariance inherent in eigendecompositions, the approach innovatively incorporates an optimal linear assignment strategy. Furthermore, it introduces an augmented monitoring scheme tailored for complex-valued time series. Experimental results on both synthetic and real-world datasets—including video and text streams—demonstrate that the proposed method matches or outperforms state-of-the-art alternatives, exhibiting strong generalization capability and practical deployment potential.

This study addresses the performance degradation of machine learning models caused by concept drift in dynamic data streams. It systematically analyzes the characteristics of concept drift and theoretically investigates, alongside empirical evaluation, the behavior of multiple learner-based detection algorithms under diverse drift scenarios—including abrupt and gradual shifts. Through comprehensive experiments on both synthetic and real-world datasets, the work compares the behavioral patterns and applicability of various detection methods, thereby deepening the understanding of underlying drift mechanisms. The findings elucidate the relative strengths and limitations of different detectors across heterogeneous environments, offering robust empirical guidance for algorithm selection in practical applications.

This study addresses the challenge of detecting and localizing mean change points in high-dimensional dependent time series, particularly under non-Gaussian distributions and temporal dependence. The authors propose an adaptive method that integrates a quadratic-form CUSUM statistic with a coordinate-wise maximum statistic to effectively capture dense and sparse changes, respectively, while employing a weighting scheme to distinguish interior from boundary change points. A key theoretical contribution lies in establishing the limiting distributions and asymptotic independence of these two statistics under non-Gaussian dependence, thereby providing rigorous justification for a Cauchy combination test. Coupled with wild binary segmentation, the approach achieves consistent estimation of multiple change points. Theoretical analysis confirms the validity of centering and scaling procedures, and extensive numerical experiments demonstrate superior detection accuracy and localization efficiency across diverse scenarios.