Integrated Time Series Research Articles

Containerized Freight Index Prediction Integrating Deep Learning and Time Series Analysis

In the context of globalization, the shipping industry, as a core pillar of international trade, is crucial for global economic stability. This study focuses on the China Containerized Freight Index (CCFI) and proposes a composite deep learning model combining CEEMDAN, CNN, LSTM, and SENet for CCFI analysis and prediction. Initially, the CEEMDAN algorithm is used to decompose the CCFI time series, extracting useful signals from noise. Then, CNN extracts features, LSTM model sequences, and SENet enhances feature representation. The CEEMDAN-CNN-LSTM-SENet model achieves superior predictive performance with MAE of 29.5171, RMSE of 33.7638, and MAPE of 1.56%. Compared to other models like ECL (MAE 77.0951, RMSE 57.3559, MAPE 3.25%), our model shows approximately 61.7% improvement in MAE. This research offers a robust tool for shipping market analysis and policy-making, providing new perspectives for handling complex economic data.

Integrating time series Sentinel-2 images and tide height to mapping tidal flats in the Chinese mainland

As a unique and important ecosystem, tidal flats provide a variety of ecosystem functions and services. Mapping tidal flats is essential for the protection and management of coastal ecosystems. However, large-scale tidal flats mapping still faces challenges due to the tidal variation and spectral similarity between tidal flats and inland wetlands. Previous methods rely on the coastlines or maximum seawater extent to exclude inland areas, which is limited by its inability to effectively differentiate tidal flats from spectrally similar inland wetlands. To address these issues, we proposed a new tidal-flat mapping method by integrating Sentinel-2 time series imagery with tide height (TH) data from ground-based tide stations on Google Earth Engine. We first generated images at the lowest and highest tidal stages, established the statistical relationship between the Normalized Difference Water Index (NDWI) of each pixel and TH, and then concatenated them into a Random Forest classifier for further classification. The statistical relationship between NDWI and TH amplified the difference between tidal flats and inland wetlands, thus significantly reducing the influence of spectral similarity. This method could produce a high-precision tidal flats map with an overall accuracy of 97.30% in the coastal zone of the Chinese mainland. By quantitatively comparing with the previous tidal flat maps, we found that the strategies of tidal-level information simulation and inland area exclusion were the two main reasons producing the differences among the maps. The proposed method does not rely on space constraints to exclude inland wetlands and can capture more estuarine tidal flats, so it can be used as a reliable means to monitor the tidal flats in large-scale areas.

The Extent of Agricultural Land Cover Change: The Case of Bautista, Pangasinan

Urban Expansion in Bautista periodically alters its land cover, significantly diminishing agricultural land in favor of built-up areas. Therefore, a comprehensive investigation of past and future land cover changes is essential for sustainable development. The main objective of this research is to determine the extent of agricultural land cover change and its spatial pattern from 2018 to 2022 and in the year 2027 in Bautista, Pangasinan. This study employed a descriptive research approach integrating time series data, spatial analysis, and simulation tools to evaluate the dynamic changes in agricultural land cover. The extent of agricultural land cover change decreased from 3,480.95 hectares in 2018 to 3,417.04 hectares in 2022, primarily near road networks and within the identified urban barangays of the municipality. The change exhibited a clustered distribution, as per the summary report from the Average Nearest Neighbor using the Geographic Information System. In contrast, built-up land cover change expanded by 60.68 hectares from 2018 to 2022, following an increasing trend. Furthermore, the projected agricultural land cover change is expected to decrease continuously from 3,417.039 hectares in 2022 to 3,406.875 hectares in 2027. The results indicate a conversion of agricultural land to built-up areas between 2018 and 2022 and a further decrease in 2027, reflecting a shift from agriculture to man-made structures. This research underscores the importance of monitoring land conversions and provides a basis for sustainable solutions to address the growing demand for urbanization without compromising the future extent of agricultural land cover in Bautista, Pangasinan.

Short-term prediction of horizontal winds in the mesosphere and lower thermosphere over coastal Peru using a hybrid model

The mesosphere and lower thermosphere (MLT) are transitional regions between the lower and upper atmosphere. The MLT dynamics can be investigated using wind measurements conducted with meteor radars. Predicting MLT winds could help forecast ionospheric parameters, which has many implications for global communications and geo-location applications. Several literature sources have developed and compared predictive models for wind speed estimation. However, in recent years, hybrid models have been developed that significantly improve the accuracy of the estimates. These integrate time series decomposition and machine learning techniques to achieve more accurate short-term predictions. This research evaluates a hybrid model that is capable of making a short-term prediction of the horizontal winds between 80 and 95 km altitudes on the coast of Peru at two locations: Lima (12°S, 77°W) and Piura (5°S, 80°W). The model takes a window of 56 data points as input (corresponding to 7 days) and predicts 16 data points as output (corresponding to 2 days). First, the missing data problem was analyzed using the Expectation Maximization algorithm (EM). Then, variational mode decomposition (VMD) separates the components that dominate the winds. Each resulting component is processed separately in a Long short-term memory (LSTM) neural network whose hyperparameters were optimized using the Optuna tool. Then, the final prediction is the sum of the predicted components. The efficiency of the hybrid model is evaluated at different altitudes using the root mean square error (RMSE) and Spearman’s correlation (r). The hybrid model performed better compared to two other models: the persistence model and the dominant harmonics model. The RMSE ranged from 10.79 to 27.04 m s−1, and the correlation ranged from 0.55 to 0.94. In addition, it is observed that the prediction quality decreases as the prediction time increases. The RMSE at the first step reached 6.04 m s−1 with a correlation of 0.99, while at the sixteenth step, the RMSE increased up to 30.84 m s−1 with a correlation of 0.5.

Trends in satellite-based ocean parameters through integrated time series decomposition and spectral analysis: 1. chlorophyll, sea surface temperature, and sea level anomaly

Abstract This study investigated trends in satellite-based chlorophyll-a (Chl-a; 1998–2022), sea surface temperature (SST; 1982–2022), and sea level anomaly (SLA; 1993–2021) from the European Space Agency’s Climate Change Initiative records, integrating time series decomposition and spectral analysis. Trends in parameters signify prolonged increases, decreases, or no changes over time. These are time series in the same space as original parameters, excluding seasonalities and noise, and can exhibit nonlinearity. Trend rates approximate the pace of change per time unit. We quantified trends using conventional linear-fit and three incrementally advancing methods for time series decomposition: simple moving average (SMA), seasonal-trend decomposition using locally estimated scatterplot smoothing (STL), and multiple STL (MSTL), across the global ocean, the Bay of Bengal, and the Chesapeake Bay. Challenges in decomposition include specifying accurate seasonal periods that are derived here by combining Fourier and Wavelet Transforms. Globally, SST and SLA trend upwards, and Chl-a has no significant change, yet regional variations are notable. We highlight the advantage of extracting multiple periods with MSTL and, more broadly, decomposition’s role in disentangling time-series components (seasonality, trend, noise) without resorting to monotonic functions, thereby preventing overlooking episodic events. Illustrations include extreme events temporarily counteracting background trends, e.g., the 2010–2011 SLA drop due to La Niña-induced rainfall over land. The continuous analysis clarifies the warming hiatus debate, affirming sustained warming. Decadal trend rates per grid cell are also mapped. These are ubiquitously significant for SST and SLA, whereas Chl-a trend rates are globally low but extreme across coasts and boundary currents.

Boosting the HP filter for trending time series with long-range dependence

This article extends recent asymptotic theory developed for the Hodrick Prescott (HP) filter and boosted HP (bHP) filter to long-range dependent time series that have fractional Brownian motion (fBM) limit processes after suitable standardization. Under general conditions, it is shown that the asymptotic form of the HP filter is a smooth curve, analogous to the finding in Phillips and Jin for integrated time series and series with deterministic drifts. Boosting the filter using the iterative procedure suggested in Phillips and Shi leads under well-defined rate conditions to a consistent estimate of the fBM limit process or the fBM limit process with an accompanying deterministic drift when that is present. A stopping criterion is used to automate the boosting algorithm, giving a data-determined method for practical implementation. The theory is illustrated in simulations and two real data examples that highlight the differences between simple HP filtering and the use of boosting. The analysis is assisted by employing a uniformly and almost surely convergent trigonometric series representation of fBM.

Deep learning model for the deformation prediction of concrete dams under multistep and multifeature inputs based on an improved autoformer

The long-term prediction of deformation in concrete dams is a critical requirement for maintaining their structural integrity over time in practical management scenarios. While most existing models predominantly address short-term and medium-term predictions, there remains a significant challenge in establishing reliable correlations within complex long-term temporal patterns and varying environmental factors. Consequently, there is limited research on multistep prediction. To solve these problems, an Autoformer-based model for multistep and multifeature dam deformation prediction is proposed. In the proposed model, the sequence decomposition of dam monitoring data is used as the basic construction module inside the depth model. Then, Deep Automatic Correlation (Deep-AutoCorrelation) mechanism is employed to obtain the long-term dependence between dam deformation and environmental load. The encoder-decoder structure integrates time series decomposition and the Deep-AutoCorrelation to predict the multistep deformation of the dam. In the multistep deformation prediction of the dam, the improved Autoformer model demonstrates state-of-the-art performance, resulting in a 48% increase in prediction accuracy across five monitoring point datasets when compared to six benchmark models. Experimental results indicate that this method effectively addresses the challenge of predicting long-term deformation of dams subjected to complex environmental loads. It demonstrates robust capabilities in extracting long-term temporal features and avoids the increase of computational complexity caused by deeper time extraction.

Interpretable and efficient RUL prediction of turbofan engines using EM-enhanced Bi-LSTM with TCN and attention mechanism

Remaining useful life (RUL) prediction for turbofan engines is important in prognostics and health management (PHM) for the maintenance and operation of critical equipment. With continuous innovations in deep learning techniques, the complexity of models continues to increase, but the interpretability and comprehensibility of the prediction results become particularly important in industrial applications. Therefore, in this study, an improved bidirectional long and short-term memory network (Bi-LSTM) based interpretable hybrid deep learning model for RUL prediction of turbofan engines is proposed, which ingeniously integrates time series convolutional networks (TCNs), expectation maximization (EM), Bi-LSTMs, and attention mechanisms. By capturing time-series features at different levels, the model adapts to the complex dynamics of turbofan engine performance evolution in an efficient and cost-effective manner. Experimental validation on the C-MAPSS dataset demonstrated that the model significantly outperforms other methods in terms of RUL prediction performance, especially in improving prediction accuracy and coping with the degradation of complex system dynamics. The largest contribution of key metrics to the model is validated through consistent results from multiple interpretable tools, providing comprehensive and consistent support for understanding and trusting prediction results in industrial applications. This study further enhances the robustness of the model and the reliability of the interpretable results by delving into the dynamic relationships between the properties of the different life stages, which not only reveal the importance of these characteristics in engine life prediction but also provide more comprehensive information about the engine performance variations by observing the dynamic relationships.

Machine-Learning model for estimating sugarcane production at crop level

Yield maps provide essential information for those who manage the field. The anticipated production data will be able to make better decisions on how resources should be used in harvesting, define market strategies and, above all, it will help evaluate treatments used on the crop. Sugar cane is the predominant crop in Huasteca Potosina, Mexico. The proposed Machine Learning model based on Random Forest Regressor integrates time series of vegetation indices extracted from Sentinel-2 images and meteorological data. The R2 and RMSE metrics (0.84 y 0.034) show the effectiveness of the model for prediction.

Optimizing User Experience: A Data-Driven Approach to Enhancing B2B Invoicing Efficiency across Multiple Platforms PT Pakar Digital Global

This study is a deep dive into applying data-driven decision-making strategies to enhance user experience and efficiency in B2B invoicing across multiple platforms at Paper.id. Given the fintech industry’s rapid evolution, maintaining a competitive edge requires constant innovation, particularly in user activation and retention within the first month of usage. This research employs a comprehensive data analysis framework, integrating Time Series Analysis, Descriptive Analysis, and Correlation Analysis methods to extract actionable insights from user data stored in MySQL databases. The study also incorporates strategic planning tools like SWOT and TOWS analysis to identify and implement UI/UX enhancements using the Agile Scrum methodology. The effectiveness of these interventions is evaluated through a robust post-analysis, comparing metrics before and after the implementation across mobile and desktop platforms. Results demonstrate a significant improvement in user activation and retention rates on mobile platforms, validating the efficacy of the applied strategies. This paper contributes valuable insights into optimizing user experience in a dynamic market, emphasizing a structured, iterative approach to achieving sustained user engagement and business growth.

Mapping high spatial resolution ionospheric total electron content by integrating Time Series InSAR with International Reference Ionosphere model

Total electron content (TEC) is a key parameter for characterizing the ionosphere. Conventional TEC measurement methods have low spatial resolution, hindering accurate representation of ionospheric spatial characteristics. Synthetic Aperture Radar (SAR) and Interferometric SAR (InSAR) have shown their potential for high-spatial-resolution TEC estimation. However, SAR-based absolute TEC estimation typically relies on full-polarimetric SAR images, while the InSAR-based method can only extract differential TEC. This study proposes a novel TEC mapping method integrating Time Series InSAR (TS-InSAR) with the International Reference Ionospheric (IRI) model and does not leverage full-polarimetric SAR data. The Advanced Land Observing Satellite-1 (ALOS-1) SAR images covering the Chile and Greenland regions were collected to test the proposed method. The TECs derived from the IRI model and International GNSS Service (IGS) and the electron density observed by the Incoherent Scattering Radar (ISR) site of SONDRESTROM were acquired to evaluate the performance and validate the accuracy of the proposed method, respectively. The results show that the proposed method can achieve TEC mapping with resolutions of tens of meters, much higher than the tens or hundreds of kilometers of the IRI and IGS TECs. The spatial features of our method-derived TECs show overall good consistency with those of the IRI and IGS TECs, but the described spatial information is more detailed. Also, compared to the original IRI model, the improvement rates of root-mean-square errors (RMSEs) between the estimated electron density and ISR observations exceeded 21.67% after the update using the TEC obtained from the proposed method. These experiments demonstrate the proposed method’s feasibility and reliability for ionospheric TEC mapping.

Contextual AI Journaling: Integrating LLM and Time Series Behavioral Sensing Technology to Promote Self-Reflection and Well-being using the MindScape App.

MindScape aims to study the benefits of integrating time series behavioral patterns (e.g., conversational engagement, sleep, location) with Large Language Models (LLMs) to create a new form of contextual AI journaling, promoting self-reflection and well-being. We argue that integrating behavioral sensing in LLMs will likely lead to a new frontier in AI. In this Late-Breaking Work paper, we discuss the MindScape contextual journal App design that uses LLMs and behavioral sensing to generate contextual and personalized journaling prompts crafted to encourage self-reflection and emotional development. We also discuss the MindScape study of college students based on a preliminary user study and our upcoming study to assess the effectiveness of contextual AI journaling in promoting better well-being on college campuses. MindScape represents a new application class that embeds behavioral intelligence in AI.

Learning spectral-indices-fused deep models for time-series land use and land cover mapping in cloud-prone areas: The case of Pearl River Delta

Mapping of highly dynamic changes in land use and land cover (LULC) can be hindered by various cloudy conditions with optical satellite images. These conditions result in discontinuities in high-temporal-density LULC mapping. In this paper, we developed an integrated time series mapping method to enhance the LULC mapping accuracy and frequency in cloud-prone areas by incorporating spectral-indices-fused deep models and time series reconstruction techniques. The proposed method first reconstructed cloud-contaminated pixels through time series filtering, during which the cloud masks initialized by a deep model were refined and updated during the reconstruction process. Then, the reconstructed time series images were fed into a spectral-indices-fused deep model trained on samples collected worldwide for classification. Finally, post-classification processing, including spatio-temporal majority filtering and time series refinement considering land–water interactions, was conducted to enhance the LULC mapping accuracy and consistency. We applied the proposed method to the cloud- and rain-prone Pearl River Delta (i.e., Guangdong–Hong Kong–Macao Greater Bay Area, GBA) and used time series Sentinel-2 images as the experimental data. The proposed method enabled seamless LULC mapping at a temporal frequency of 2–5 days, and the production of 10 m resolution annual LULC products in the GBA. The assessment yielded a mean overall accuracy of 87.01% for annual mapping in the four consecutive years of 2019–2022 and outperformed existing mainstream LULC products, including ESA WorldCover (83.98%), Esri Land Cover (85.26%), and Google Dynamic World (85.06%). Our assessment also reveals significant variations in LULC mapping accuracies with different cloud masks, thus underscoring their critical role in time series LULC mapping. The proposed method has the potential to generate seamless and near real-time maps for other regions in the world by using deep models trained on datasets collected globally. This method can provide high-quality LULC data sets at different time intervals for various land and water dynamics in cloud- and rain-prone regions. Notwithstanding the difficulties of obtaining high-quality LULC maps in cloud-prone areas, this paper provides a novel approach for the mapping of LULC dynamics and the provision of reliable annual LULC products.

Comparative analysis of different approaches to the Ukrainian residential property market evolution modelling and its forecast for the years 2019–2024

PurposeThis study aims to provide a comprehensive analysis of various approaches to the residential property market evolution modelling and to examine the macroeconomic fundamentals that have shaped this market development in Ukraine in recent years.Design/methodology/approachThe study uses a comprehensive data set encompassing relevant macroeconomic indicators and historical apartment prices. Multifactor linear regression (MLR) and ridge regression (RR) models are constructed to identify the impact of multiple predictors on apartment prices. Additionally, the ARIMAX model integrates time series analysis and external factors to enhance modelling and forecasting accuracy.FindingsThe investigation reveals that MLR and RR yield accurate predictions by considering a range of influential variables. The hybrid ARIMAX model further enhances predictive performance by fusing external indicators with time series analysis. These findings underscore the effectiveness of a multidimensional approach in capturing the complexity of housing price dynamics.Originality/valueThis research contributes to the real estate modelling and forecasting literature by providing an analysis of multiple linear regression, RR and ARIMAX models within the specific context of property price prediction in the turbulent Ukrainian real estate market. This comprehensive analysis not only offers insights into the performance of these methodologies but also explores their adaptability and robustness in a market characterized by evolving dynamics, including the significant influence of external geopolitical factors.

An efficient load forecasting technique by using Holt‐Winters and Prophet algorithms to mitigate the impact on power consumption in COVID‐19

AbstractIt is strongly recommended to implement effective long‐term load forecasting for future power generation in the new architecture of the smart grid and buildings. This method is essential for the smart grid's stability, power demand estimation, and an improved energy management system, which will enhance integration between efficient demand response and distributed renewable energy sources. However, due to influencing elements including climatic, societal, and seasonal aspects, it is quite challenging to perform energy prediction with high accuracy. To estimate the load demand before and during the time period of the COVID‐19 paradigm with its diversity and complexity, the authors present and integrate time series forecasting techniques such as Holt‐Winters and Prophet algorithms. In comparison to the Holt‐Winters model, the Prophet model has shown to be more noise‐resistant. Additionally, the Prophet model surpasses the Holt‐Winters model according to the generalisability test of the two models by using the hourly driven power consumption data from Houston, Texas, USA. The resultant constraints and influential factors are discussed, and experimental results can be validated from the pivotal outcome.

Impacts of pandemic service adaptations on job accessibility: A case study of the Bay Area Rapid Transit

The COVID-19 pandemic highlighted transit's crucial role as a social service, ensuring access to essential destinations. Despite this, unprecedented ridership lows forced agencies to implement service cuts, disproportionately affecting essential workers and vulnerable populations. However, the full extent of these impacts remains underexplored. While existing literature examines transit agency responses during the pandemic, much of the focus has been on public health and safety measures, overlooking the specifics of service adjustment strategies implemented. This study contributes to our understanding of transit agency pandemic responses throughout the pre-, peak-, and post-pandemic phases by 1) characterizing patterns in transit service adjustments and 2) extending pandemic accessibility literature by examining job-specific impacts. The framework integrates time series clustering, qualitative review of agency press releases, and transit accessibility analysis, using only publicly available data. Through a case study of Bay Area Rapid Transit, we find distinct clusters of stations characterized by patterns in weekday morning peak service restoration and station area demographics. While impacts to accessibility varied by time of day, the relative ordering of accessibility levels across income and race/ethnicity remained consistent throughout the pandemic. These findings contribute to our understanding of service adaptation impacts and inform equitable response strategies for future service planning and disruptions.

Anomaly Detection in Species Distribution Patterns: A Spatio-Temporal Approach for Biodiversity Conservation

Monitoring and maintaining biodiversity are crucial for the ecological balance and overall health of ecosystems. Detecting anomalies in species distribution patterns is essential for identifying areas of concern and implementing timely conservation measures. This paper introduces an approach to detect anomalies in spatiotemporal species activity data by integrating time series analysis, machine learning techniques, and spatial statistics. Our method identifies distribution patterns of various species and groups regions with similar patterns, enabling the segmentation of the study area into distinct categories. Within these categorized regions, we apply a combination of clustering algorithms and outlier detection techniques to pinpoint anomalous behaviors in species distribution. In order to confirm the reliability of our findings, we cross-reference them with verified data acquired through field observations or other credible data sources. These corroborations indicate that anomalies are frequently indicative of sudden variations in species numbers or unexpected alterations in spatial distribution at certain places and times. To gain a more robust understanding of how species are distributed, we curate a data set that excludes these anomalous observations and use it to develop a predictive algorithm. Our analysis shows that a predictive model trained on this refined, anomaly-free dataset achieves a lower normalized mean square error (NMSE), which suggests a higher level of predictive accuracy as compared to a model trained on data containing anomalies. Utilizing this methodology can facilitate the creation of effective conservation plans and contribute to more sustainable ecosystem management.

An Integrated Time-Series Relative Soil Moisture Monitoring Method Based on a SAR Backscattering Model

An Integrated Time-Series Relative Soil Moisture Monitoring Method Based on a SAR Backscattering Model

The Research on Pricing and Replenishment Optimization of Fresh Supermarket Vegetable Products based on Sales Data

This study delves into the challenges and complexities of vegetable sales management in modern commercial environments, particularly in fresh food supermarkets. By utilizing descriptive statistics and visual analysis of sales data, the study reveals the correlations between different vegetable categories and quantifies these correlations using Pearson's correlation coefficient. Subsequently, by integrating time series analysis and multi-objective programming, a mathematical model is constructed, aimed at maximizing profits under specific constraints. The innovation of this research lies in its comprehensive consideration of category-level sales management and the application of modern optimization algorithms for replenishment and pricing strategies. The uniqueness of this paper is in its integrative approach to the vegetable sales problem, providing refined mathematical models and advanced optimization methods. Finally, the paper thoroughly describes the steps of the research design, including sales data analysis, cost markup analysis, and the construction of an optimization model based on time series analysis and multi-objective programming, intending to offer supermarkets a vegetable sales management plan adaptable to market changes.

A Supply Chain-Oriented Model to Predict Crude Oil Import Prices in South Korea Based on the Hybrid Approach

Although numerous studies have explored key variables for forecasting crude oil prices, the role of supply chain factors has often been overlooked. In the face of global risks such as COVID-19, the Russia–Ukraine war, and the U.S.–China trade dispute, supply chain management (SCM) has evolved beyond an individual company’s concern. This research investigates the impact of a supply chain-oriented variable on the forecasting of crude oil import prices in South Korea. Our findings reveal that models incorporating the Global Supply Chain Pressure Index (GSCPI) outperform those without it, emphasizing the importance of monitoring supply chain-related variables for stabilizing domestic prices for policy makers. Additionally, we propose a novel hybrid factor-based approach that integrates time series and machine learning models to enhance the prediction performance of oil prices. This endeavor is poised to serve as a foundational step toward developing methodologically sound forecasting models for oil prices, offering valuable insights for policymakers.