Change Point Detection Research Articles

Change Point Detection in Length-Biased Weibull Distribution for Random Censored Data Based on Modified Information Criterion

Change Point Detection in Length-Biased Weibull Distribution for Random Censored Data Based on Modified Information Criterion

Some clustering-based change-point detection methods applicable to high dimension, low sample size data

Detection of change-points in a sequence of high dimensional observations is a challenging problem, and this becomes even more challenging when the sample size (i.e., the sequence length) is small. In this article, we propose some change-point detection methods based on clustering, which can be conveniently used in such high dimension, low sample size situations. First, we consider the single change-point problem. Using k-means clustering based on a suitable dissimilarity measures, we propose some methods for testing the existence of a change-point and estimating its location. High dimensional behavior of these proposed methods are investigated under appropriate regularity conditions. Next, we extend our methods for detection of multiple change-points. We carry out extensive numerical studies and analyze a real data set to compare the performance of our proposed methods with some state-of-the-art methods.

Uncovering insights from big data: change point detection of classroom engagement

Expectations of big data across various fields, including education, are increasing. However, uncovering valuable insights from big data is like locating a needle in a haystack, and it is difficult for teachers to use educational big data on their own. This study aimed to understand changes in student participation rates during classes and teachers’ teaching styles by analyzing educational big data. In the analysis, data from 120 students and two mathematics class teachers at a public junior high school in Japan were used. We applied the pruned exact linear time (PELT) algorithm to automatically identify significant changes in student participation during class. Based on the information obtained, we analyzed the interaction logs of teachers’ e-book readers and clarified the relationship between student participation rates and teacher behavior patterns. Change point detection using the PELT algorithm showed a high F1-score of 0.7929, indicating good overall performance. We also investigated whether there was a relationship between class differences and teachers’ actions and found a statistically significant difference. The results provide clues for improving student learning engagement and teachers’ teaching styles, and they are expected to improve the quality of education by automatically identifying notable cases from educational big data. However, further research is required to improve data analysis methods, such as adjusting the parameters of algorithms based on the situation.

Change point detection of events in molecular simulations using dupin

Particle tracking is commonly used to study time-dependent behavior in many different types of physical and chemical systems involving constituents that span many length scales, including atoms, molecules, nanoparticles, granular particles, and even larger objects. Behaviors of interest studied using particle tracking information include disorder-order transitions, thermodynamic phase transitions, structural transitions, protein folding, crystallization, gelation, swarming, avalanches and fracture. A common challenge in studies of these systems involves change detection. Change point detection discerns when a temporal signal undergoes a change in distribution. These changes can be local or global, instantaneous or prolonged, obvious or subtle. Moreover, system-wide changes marking an interesting physical or chemical phenomenon (e.g. crystallization of a liquid) are often preceded by events (e.g. pre-nucleation clusters) that are localized and can occur anywhere at anytime in the system. For these reasons, detecting events in particle trajectories generated by molecular simulation is challenging and typically accomplished via ad hoc solutions unique to the behavior and system under study. Consequently, methods for event detection lack generality, and those used in one field are not easily used by scientists in other fields. Here we present a new Python-based tool, dupin, that allows for universal event detection from particle trajectory data irrespective of the system details. dupin works by creating a signal representing the simulation and partitioning the signal based on events (changes within the trajectory). This approach allows for studies where manual annotating of event boundaries would require a prohibitive amount of time. Furthermore, dupin can serve as a tool in automated and reproducible workflows. We demonstrate the application of dupin using three examples and discuss its applicability to a wider class of problems. Program summaryProgram Title:dupinCPC Library link to program files:https://doi.org/10.17632/kjcn97zc46.1%Developer's repository link::https://github.com/glotzerlab/dupinLicensing provisions: BSD 3-clauseProgramming language: PythonNature of problem: In the field of molecular simulations, detecting structural transitions or events within trajectories can be both challenging and time-consuming for larger studies due to the requirement of a manual approach. This issue is particularly pronounced in studies involving hundreds or thousands of simulations, where manual detection and analysis of transitions become infeasible. Our goal is to develop an automated, accurate and efficient method for detecting transition points in simulation trajectories, which both saves time and aids researchers in uncovering important events and their underlying causes in various systems. Additionally, we aim to facilitate new machine learning applications to important materials problems such as predicting and designing crystallization pathways, predicting defect formation, and describing the behavior of active matter, all of which involve structural transitions occurring over time. The developed method should be applicable to offline and online detection, enabling event-dependent triggers for advanced simulation/experimental protocols and efficient processing and storing of data.Solution method: We develop a versatile python package called dupin for detecting molecular events and structural transitions in simulation trajectories. dupin's workflow pipeline includes three major stages: data preprocessing, data augmentation, and detection. The components of this pipeline collectively improve the accuracy and efficiency of identifying structural changes in particle trajectory data. In data preprocessing, we generate and aggregate data into a comprehensive representation of the system. Data augmentation techniques such as feature selection and dimensionality reduction counteract the noise arising from high-dimensional data and enhance computational performance. We detect change points within the trajectory indicating transition events using a cost-based event detection method. In dupin, we implement two cost functions based on piecewise linear fits, which offer different levels of sensitivity to sudden shifts and changes in the signal. The package can use any cost-based detection algorithm but has a special interface for the Python package ruptures. Regardless of detection algorithm, we use the cost function and “elbow” detection to determine the correct number of change points. The detection scheme can be applied both offline and online, enabling real-time analysis of molecular events as simulations progress. As an example, dupin may be used to trigger a high frequency storage of frames within a simulation upon nucleation and subsequent solidification of a liquid into a crystal. Our method demonstrates a high degree of accuracy in detecting transition points within simulation trajectories when provided with informative descriptors. By automating the detection process, our solution enables efficient change point detection for studies with large-scale simulations.Additional comments including restrictions and unusual features: Our package, dupin, has great promise in detecting transition points within simulation trajectories with a high degree of accuracy; nonetheless, it is essential to note that it relies heavily on the selection of informative descriptors. The accuracy of the detection may be compromised if the chosen descriptors do not effectively capture the changes in a system's properties. However, this restriction can be mitigated by selecting a diverse range of descriptors and applying a feature selection tool to refine the signal.

Change point detection of health stage and remaining useful lifetime prediction for electro-mechanical brake unit on trains

Electro-mechanical brake units (EMBUs) have complex degradation process. There are challenges such as component fault coupling, unclear degradation characteristics, random disturbance and so on. A remaining useful life (RUL) framework for EMBUs is developed in this paper. To eliminate the issue of time lag, a data pre-processing method based on zero-phase filtering is proposed for health indicator (HI) construction. Then, the 3 health stages (HSs) characteristics of degradation process is proposed, and is modeled using the Wiener process. To overcome perturbation of large diffusion coefficient and random fluctuation, the change point (CP) detection methods based on the variance ratio hypothesis test and maximal overlap discrete wavelet transform — multi-resolution analysis (MODWT-MRA) are developed. Besides, a consecutive anomaly window detection strategy using sliding windows is employed. Finally, the MODWT-MRA-based RUL prediction methods is proposed. The proposed methods are verified by the durability test data. The proposed data pre-processing method can avoid 10% time error compared to the common difference method. The proposed MODWT-MRA-based CP detection and RUL prediction method can overcome the interference of large random fluctuations in the non-healthy state of EMBU.

Change-point detection using diffusion maps for sleep apnea monitoring with contact-free sensors

Change-point detection using diffusion maps for sleep apnea monitoring with contact-free sensors

Bayesian neuromorphic imaging for single-photon LiDAR.

This paper proposes a Bayesian approach to enable single photon avalanche diode (SPAD) arrays to be used as pseudo event cameras that report changes in the scene. Motivated by the working principle of event cameras, which produce sparse events associated with light flux changes, we adopt a changepoint detection strategy to generate intensity and depth change event streams from direct time-of-flight (dToF) sequences measured by SPAD arrays. Although not our main goal, the algorithm also produces as a by-product, intensity and depth estimates. Unlike the output of passive event cameras that only correspond to light flux changes, the change events detected from the sequential dToFs can relate to changes in light flux and/or depth. The integration of the proposed Bayesian approach with single-photon LiDAR (SPL) systems provides a novel solution to achieve active neuromorphic 3D imaging that offers the advantages of significantly reduced output redundancy and in particular the capacity to report scene depth changes. For each pixel of the SPAD array, asynchronous events are generated by performing online Bayesian inference to detect changepoints and estimate the model parameters simultaneously from individual single-photon measurements. Experiments are conducted on synthetic data and real dToF measurements acquired by a 172×126 pixel SPAD camera to demonstrate the feasibility and efficiency of the proposed Bayesian approach.

Research on Match Trends in Tennis Competitions Based on Bayesian Online Change Point Detection and Random Forest Prediction Model

Research on Match Trends in Tennis Competitions Based on Bayesian Online Change Point Detection and Random Forest Prediction Model

Spatiotemporal analysis and forecasting of PV systems, battery storage, and EV charging diffusion in California: A graph network approach

Understanding the dynamics of clean energy adoption is crucial for shaping effective energy policies and strategies. To understand the spatiotemporal variability and bottom-up spontaneity of the diffusion of clean energy technologies, this study examined the spatiotemporal diffusion of Photovoltaic (PV) systems, Battery Energy Storage Systems (BESS), and Electric Vehicle Charging Stations (EVCS) across 1773 postcode areas in California over the past 35 years. The Bayesian change point detection algorithm (BCDA) was initially utilized to identify temporal change points in the evolution of these technologies, revealing their diffusion trends over time. Then, the Louvain community detection algorithm (LCDA) was employed to elucidate spatial diffusion patterns across regions, offering insights into the geographic proliferation of clean energy solutions. Furthermore, to forecast future distributions, the spatiotemporal graph convolutional network (STGCN) model was applied, adeptly capturing the multi-stage, nonlinear characteristics of clean energy diffusion marked by significant spatiotemporal interactions. With historical data and spatial proximity, the STGCN model demonstrated fine forecasting precision (R^2 ≥ 0.78). This study validates the effectiveness of a graph network-based approach for analyzing the diffusion of clean energy technologies. It recommends policies tailored to account for technological maturity, geographic disparities, and diffusion stages, aiming for equitable and sustainable clean energy development in the region.

Social distancing during the COVID-19 pandemic: Potential impact and correlation with asthma

BackgroundNon-pharmaceutical interventions have been implemented globally to control the COVID-19 pandemic and have been shown to alleviate both allergies and respiratory infections. Although mask-wearing is an accepted non-pharmaceutical intervention, the effects of social distancing have not been thoroughly evaluated. ObjectivesTo evaluate the effects of social distancing on asthma trends in Seoul, South Korea. MethodsThis study included data from the National Health Insurance Service of South Korea, covering approximately 10 million people in Seoul. Daily and monthly data of patients with asthma from 2018 to 2021 were examined, and the degree of social distancing performance was measured using the number of subway users as an index. Pearson's correlation coefficient was used to determine the relationship between the two indices. The change-point detection technique, cross-correlation, and Granger causality method were used to assess the temporal causality between social distancing and asthma. ResultsThe number of patients with asthma decreased by 42.4 % from 2019 to 2020, while that of subway users decreased by 26.3 % during this period. Pearson's correlation analysis revealed significant positive correlations. Asthma and subway users showed a significant change in incidence following the implementation of social distancing; subway users showed a causal relationship with patients with asthma. ConclusionOur results showed that the number of subway users decreased after the implementation of strict social distancing, coinciding with a decrease in the number of patients with asthma. These findings suggest that social distancing measures implemented to control COVID-19 may reduce the incidence and exacerbation of asthma.

High-dimensional detection of Landscape Dynamics: a Landsat time series-based algorithm for forest disturbance mapping and beyond

ABSTRACT Time series analysis of medium-resolution multispectral satellite imagery is critical to investigate forest disturbance dynamics at the landscape scale. In particular, the spatial, temporal, and radiometric consistency of Landsat time series data provides unprecedented insight into past disturbances that occurred over the last four decades. Several Landsat time series-based algorithms have been developed to automate the detection of forest disturbances. However, automated detection of non-stand-replacing disturbances based on Landsat time series remains a challenging task due to the difficulty of effectively separating them from spectral noise. Here, we present the High-dimensional detection of Landscape Dynamics (HILANDYN) algorithm, which exploits spatial and spectral information provided by Landsat time series to detect forest disturbance dynamics retrospectively. A novel and unsupervised procedure for changepoint detection in high-dimensional time series allows HILANDYN to perform the temporal segmentation of inter-annual time series into linear trends. The algorithm embeds a noise filter to remove spurious changepoints caused by residual spectral noise in the time series. We tested HILANDYN to detect disturbances that occurred in the forests of the European Alps over a period of 39 years, i.e. between 1984 and 2022, and evaluated its accuracy using a validation dataset of 3000 plots randomly located inside and outside the disturbed patches. We compared HILANDYN with the Bayesian Estimator of Abrupt change, Seasonality, and Trend (BEAST), which is a well-established and high-performing time series-based algorithm for changepoint detection. The quantitative results highlighted that the number of bands, i.e. original Landsat bands and spectral indices, included in the high-dimensional time series and the threshold controlling the significance of changepoints strongly influenced the user’s accuracy (UA). Conversely, changes in the combinations of bands primarily affected the producer’s accuracy (PA). HILANDYN achieved an F1 score of 0.801, which increased to 0.833 when we activated the noise filter, allowing the algorithm to balance UA (83.1%) and PA (83.5%). The qualitative results showed that disturbed forest patches detected by HILANDYN were characterized by a high spatio-temporal consistency, regardless of the disturbance severity. Furthermore, our algorithm was able to detect forest patches associated with secondary disturbances, such as salvage logging, that occur in close succession with respect to the primary event. The comparison with BEAST evidenced a similar sensitivity of the algorithms to non-stand-replacing events, as both achieved comparable PA. However, BEAST struggled to balance UA and PA when using a single parameter set, achieving a maximum F1 score of 0.717. Moreover, the computational efficiency of BEAST in processing high-dimensional time series was very limited due to its univariate nature based on the Bayesian approach. The adaptability of HILANDYN to detect a wide range of disturbance severities using a single parameter set and its computational efficiency in handling high-dimensional time series promotes its scalability to large study areas characterized by heterogeneous ecological conditions.

Change Point Detection with Adaptive Measurement Schedules for Network Performance Verification

When verifying that a communications network fulfills its specified performance, it is critical to note sudden shifts in network behavior as quickly as possible. Change point detection methods can be useful in this endeavor, but classical methods rely on measuring with a fixed measurement period, which can often be suboptimal in terms of measurement costs. In this paper, we extend the existing framework of change point detection with a notion of physical time. Instead of merely deciding when to stop, agents must now also decide at which future time to take the next measurement. Agents must now minimize the necessary number of measurements pre- and post-change, while maintaining a trade-off between post-change delay and false alarm rate. We establish, through this framework, the suboptimality of typical periodic measurements and propose a simple alternative, called crisis mode agents. We show analytically that crisis mode agents significantly outperform periodic measurements schemes. We further verify this in numerical evaluation, both on an array of synthetic change point detection problems as well as on the problem of detecting traffic load changes in a 5G test bed through end-to-end RTT measurements.

Controlled Discovery and Localization of Signals via Bayesian Linear Programming

Scientists often must simultaneously localize and discover signals. For instance, in genetic fine-mapping, high correlations between nearby genetic variants make it hard to identify the exact locations of causal variants. So the statistical task is to output as many disjoint regions containing a signal as possible, each as small as possible, while controlling false positives. Similar problems arise, for example, when locating stars in astronomical surveys and in changepoint detection. Common Bayesian approaches to these problems involve computing a posterior distribution over signal locations. However, existing procedures to translate these posteriors into credible regions for the signals fail to capture all the information in the posterior, leading to lower power and (sometimes) inflated false discoveries. We introduce Bayesian Linear Programming (BLiP), which can efficiently convert any posterior distribution over signals into credible regions for signals. BLiP overcomes an extremely high-dimensional and nonconvex problem to verifiably nearly maximize expected power while controlling false positives. Applying BLiP to existing state-of-the-art analyses of UK Biobank data (for genetic fine-mapping) and the Sloan Digital Sky Survey (for astronomical point source detection) increased power by 30%–120% in just a few minutes of additional computation. BLiP is implemented in pyblip (Python) and blipr (R). Supplementary materials for this article are available online, including a standardized description of the materials available for reproducing the work.

Robust change-point detection for functional time series based on U-statistics and dependent wild bootstrap

The aim of this paper is to develop a change-point test for functional time series that uses the full functional information and is less sensitive to outliers compared to the classical CUSUM test. For this aim, the Wilcoxon two-sample test is generalized to functional data. To obtain the asymptotic distribution of the test statistic, we prove a limit theorem for a process of U-statistics with values in a Hilbert space under weak dependence. Critical values can be obtained by a newly developed version of the dependent wild bootstrap for non-degenerate 2-sample U-statistics.

Data-driven wind turbine sensor health validation

This paper presents a data-driven approach for detecting anomalies in wind turbine sensors, specifically anemometers and wind vanes, and the development of a smart alerting system. The study focuses on utilizing SCADA and reanalysis (ERA5) data for accurate anomaly detection and reducing false alarms through smart change point detection algorithms. The methodology involves modeling normal behavior, detecting change points, and comparing power curves before and after these points. For anemometer anomaly detection, a three-year SCADA dataset from an offshore wind farm and a synthetic dataset is used, employing an XGBoost model and the PELT algorithm for change point detection. Wind vane anomalies are identified using a nine month dataset from seven turbines, with synthetic alterations to simulate misalignments. Results show successful detection of sudden changes in wind speed and direction, with smart alarms assisting operators in decision-making. This research enhances wind turbine condition monitoring, improving reliability and efficiency.

Change Point Detection for Automatic Time Series Forecasting

Time series analysis, the process of learning patterns in time series data and making future predictions, is a challenging machine learning endeavor: the intricate and error-prone process of manual data fitting limits efficiency and scalability to larger datasets. Our research builds upon structure discovery research, paving the way for more effective short-term forecasting. We combine the adaptability of automatic change point detection with a fixed kernel composition, achieving accuracy comparable to traditional manual methods while reducing analysis time. This hybrid approach offers the best of both worlds: leveraging human expertise for precise fitting and capturing specific trends while utilizing an automated technique to recognize shifts and model complex relationships within the data dynamically. By demonstrating the effectiveness of automatic change point detection in conjunction with kernel composition, we work to develop time series usage in data analysis.

Denoising Ambulatory Electrocardiogram Signal Using Interval Dedependent Thresholds based Stationary Wavelet Transform

Noise contamination in electrocardiogram (ECG) monitoring systems can lead to errors in analysis and diagnosis, resulting in a high false alarm rate (FAR). Various studies have been conducted to reduce or eliminate noise in ECG signals. However, some noise characteristics overlap with the frequency range of ECG signals, which occur randomly and are transient. This results in shape alteration and amplitude reduction in P and R waves. The author proposed a framework for eliminating noise in ECG signals using the stationary wavelet transform method and interval-dependent thresholds (IDT) based on the change point detection method to address these challenges. The proposed framework decomposes the input electrocardiogram (ECG) signal at a specific level using the Stationary Wavelet Transform method, resulting in detail and approximation coefficients. Interval detection focuses on the initial detailed coefficient, d1, chosen due to its significant content of noise coefficients, especially high-frequency noise. Subsequently, threshold values are computed for each interval. Hard and soft thresholding processes are then applied individually to each interval. Finally, reconstruction occurs using the inverse stationary wavelet transform method on the threshold coefficient outcomes. Two measurement matrices, root mean square error (RMSE) and percentage root mean squared difference (PRD), were used to measure the performance of the proposed framework. In addition, the proposed framework was compared to stationary wavelet transform (SWT) and discrete wavelet transform (DWT). The test results showed that the proposed method outperforms DWT and SWT. The proposed framework obtained an average increase in RMSE scores of 18% and 45% compared to the SWT and DWT methods, respectively, and PRD values of 17% and 37% compared to the SWT and DWT methods, respectively. So, using IDT in the stationary wavelet transform method can improve the denoising performance. With the development of this new framework for denoising ECG signals, we hope it can become an alternative method for other researchers to utilize in denoising ECG signals.

A dual-model adaptive Kalman filtering for remaining useful life prediction method based on feature fusion and online TSP recognition

Remaining Useful Life (RUL) prediction is of significant importance for the safe operation of machines. However, the Health Indicators (HIs) constructed by a single feature of machines cannot precisely describe the non-stationary and accelerating degradation operation. Moreover, poor identification of Time to Start Prediction (TSP), and the limited adaptability of models restrict the accuracy of RUL prediction. A Dual-Model Adaptive Kalman Filter (DMAKF) RUL prediction method based on feature fusion and online TSP recognition is proposed. Firstly, a feature-fusion strategy is proposed, integrating various information to construct distinguishing HIs. Secondly, a statistic-based change-point detection approach is investigated to online recognize TSP, balancing the sensitivity and interference immunity. Finally, the DMAKF is conducted for adaptive adjustment of RUL prediction in real-time. The proposed method is validated on two sets of bearing datasets from different sources. Comparison with four other methods demonstrates the effectiveness of the proposed method.

Analysis of trends and abrupt changes in groundwater and meteorological droughts in the United Kingdom

In recent years, the escalating impact of climate change, coupled with growing concerns regarding groundwater drought, has underscored the critical importance of analyzing hydrological trends. This is a foundational step in advancing effective water resource planning and management. This study presents a regional investigation of groundwater and meteorological droughts in the UK, utilizing the Standardized Groundwater Index (SGI) and the Standardized Precipitation Index (SPI). The analysis encompasses drought duration and an evaluation of the correlation between the SGI and SPI. Subsequently, the seasonal Mann-Kendall (MK) test and the Bayesian Changepoint Detection and Time Series Decomposition (BEAST) algorithm were employed to assess overall trends for both SGI and SPI. Additionally, these methods were utilized to pinpoint trend and seasonal abrupt changes within the time series of both indices.The seasonal MK test revealed statistically significant trends in many SGI stations across the UK, spanning from Northern Scotland down to the South of England, encompassing both increasing and decreasing trends. Additionally, the MK test showed slightly positive trends at the low SPI time scale, though these were not statistically significant. In contrast, more consistent and significant increasing trends were observed throughout the entire UK, except for Northern Scotland, which exhibited negative trends. Despite indications of a progressively more humid UK, the BEAST analysis enabled the detection of decreasing abrupt changes in trends. These changes result in sudden transitions from wet to dry conditions, which can be challenging to identify using the MK test alone. However, BEAST revealed both positive and negative abrupt changes across the entire UK, as well as positive or negative variations in seasonality. This suggests a high degree of variability in seasonal patterns. In summary, this study provides a detailed assessment of groundwater and meteorological drought trends and abrupt changes, considering the impact of climate change on various regions of the UK.

Elevating Cryptocurrency Predictions: Bidirectional LSTM Methodology

The system proposed in this paper aims to predict cryptocurrency prices using Bi-Directional Long Short- Term Memory (LSTM), leveraging historical data obtained from Yahoo Finance and CoinGecko APIs. The goal is to assess LSTM models effectiveness in forecasting cryptocurrency prices and offer an interactive interface for users to visualize historical and forecasted prices. Several research works have been conducted on the prediction of cryptocurrency prices through various Deep Learning (DL) based algorithms. This project comprises two main approaches : one involves data analysis, LSTM modeling, and change point detection using Yahoo Finance data, while the other focuses on LSTM model training and price prediction using CoinGecko API data. The paper suggests that the prediction models it presents are useful for traders, investors, [6] and finance academics and are close to accurate at predicting the values of cryptocurrencies. Future research will examine more advanced deep learning architectures, primarily Transformer-based models like the GPT series, to improve pattern detection in bitcoin data. Integrating other data sources, such as sentiment analysis or blockchain measurements, may increase the accuracy of forecasting. With further research into cutting-edge techniques, cryptocurrency forecasting will get better and provide stakeholders with more information to help them make informed decisions. Keywords— Cryptocurrency ; forecasting ; Bi-Directional LSTM Model ; Time-series forecasting ; Machine learning