Search Results (773)

Corner radius end mills (CREMs) are widely used in machining due to their unique tool geometry, which improves surface quality. Variations in cutting force during machining significantly impact machining quality. Therefore, precisely predicting cutting forces is critical for controlling machining chatter and enhancing accuracy. Traditional element force models have complex formulas and high computational demands when considering tool runout. This paper proposes a hybrid prediction model for CREMs that integrates the separate-edge-forecast method and the BP neural network. The integration approach incorporates runout effects into cutting force coefficients and addresses nonlinear effects from runout. The accuracy of the cutting force prediction model was validated through side milling on 7075 aluminum alloy. The results indicate that the maximum error between the predicted and measured forces is 9.43%, demonstrating that this model ensures high prediction accuracy while reducing computation cost. Full article

Time series clustering plays a vital role in various analytical and pattern recognition tasks by partitioning structurally similar sequences into semantically coherent groups, thereby facilitating downstream analysis. However, building high-quality clustering models remains challenging due to three key issues: (i) capturing dynamic shape variations across sequences, (ii) ensuring discriminative cluster structures, and (iii) enabling end-to-end optimization. To address these challenges, we propose KDiscShapeNet, a structure-aware clustering framework that systematically extends the classical k-Shape model. First, to enhance temporal structure modeling, we adopt Kolmogorov–Arnold Networks (KAN) as the encoder, which leverages high-order functional representations to effectively capture elastic distortions and multi-scale shape features of time series. Second, to improve intra-cluster compactness and inter-cluster separability, we incorporate a dual-loss constraint by combining Center Loss and Supervised Contrastive Loss, thus enhancing the discriminative structure of the embedding space. Third, to overcome the non-differentiability of traditional K-Shape clustering, we introduce Differentiable k-Shape, embedding the normalized cross-correlation (NCC) metric into a differentiable framework that enables joint training of the encoder and the clustering module. We evaluate KDiscShapeNet on nine benchmark datasets from the UCR Archive and the ETT suite, spanning healthcare, industrial monitoring, energy forecasting, and astronomy. On the Trace dataset, it achieves an ARI of 0.916, NMI of 0.927, and Silhouette score of 0.931; on the large-scale ETTh1 dataset, it improves ARI by 5.8% and NMI by 17.4% over the best baseline. Statistical tests confirm the significance of these improvements (p < 0.01). Overall, the results highlight the robustness and practical utility of KDiscShapeNet, offering a novel and interpretable framework for time series clustering. Full article

The article presents an analysis of current (during the period 1985–2022) and projected (during the period 2025–2099) changes in the hydrological regime of the Buktyrma, Yesil, and Zhaiyk river basins in Kazakhstan under the conditions of global climate change. This study is based on the integration of data from General Circulation Models (GCMs) of the sixth phase of the CMIP6 project, socio-economic development scenarios SSP2-4.5 and SSP5-8.5, as well as the results of hydrological modelling using the SWIM model. The studies were carried out with an integrated approach to hydrological change assessment, taking into account scenario modelling, uncertainty analysis and the use of bias correction methods for climate data. A calculation method was used to analyse the intra-annual distribution of runoff, taking into account climate change. Detailed forecasts of changes in runoff and intra-annual water distribution up to the end of the 21st century for key water bodies in Kazakhstan were obtained. While the projections of river flow and hydrological parameters under CMIP6 scenarios are actively pursued worldwide, few studies have explicitly focused on forecasting intra-annual flow distribution in Central Asia, calculated using a methodology appropriate for this region and using CMIP6 ensemble scenarios. There have been studies on changes in the intra-annual distribution of runoff for individual river basins or local areas, but for the historical period, there have also been studies on modelling runoff forecasts using CMIP6 climate models, but have been very few systematic publications on the distribution of predicted intra-annual runoff in Central Asia, and this issue has not been fully studied. The projections suggest an intensification of flow seasonality (1), earlier flood peaks (2), reduced summer discharges (3) and an increased likelihood of extreme hydrological events under future climatic conditions. Changes in the seasonal structure of river flow in Central Asia are caused by both climatic factors—temperature, precipitation and glacier degradation—and significant anthropogenic influences, including irrigation and water management structures. These changes directly affect the risks of flooding and water shortages, as well as the adaptive capacity of water management systems. Given the high level of water management challenges and interregional conflicts over water use, the intra-annual distribution of runoff is important for long-term planning, the development of adaptation measures, and the formulation of public policy on sustainable water management in the face of growing climate challenges. This is critically important for water, agricultural, energy, and environmental planning in a region that already faces annual water management challenges and conflicts due to the uneven seasonal distribution of resources. Full article

There is large scale data collected from the various Wi-Fi networks on modern university campuses which contribute to observing student behavioral patterns. This paper explores the use of Wi-Fi connection information and internet browsing habits to forecast student dining preferences, improving data-driven models for campus eating service optimizations. This study combines spatial–temporal features with browsing behavior analysis and employs advanced machine learning techniques to develop a multi-modal learning framework. Moreover, when Chinese consumers go out to eat, the analysis of anonymized Wi-Fi data also reveals considerable relationships among digital footprints and dining choices using a predictive model that can reach an accuracy level between 84 and 88%. The discoveries assist in the advancement of educational data mining and are beneficial for the real-world optimization of campus services, all under strong privacy protection using an end-to-end comprehensive data protection framework. Full article

To tackle the limitations in simultaneously modeling long-term dependencies in the time dimension and nonlinear interactions in the feature dimension, as well as their inability to fully reflect the impact of real-time load changes on spatial dependencies, a short-term multi-energy load forecasting method based on Transformer Spatio-Temporal Graph neural network (TSTG) is proposed. This method employs a multi-head spatio-temporal attention module to model long-term dependencies in the time dimension and nonlinear interactions in the feature dimension in parallel across multiple subspaces. Additionally, a dynamic adaptive graph convolution module is designed to construct adaptive adjacency matrices by combining physical topology and feature similarity, dynamically adjusting node connection weights based on real-time load characteristics to more accurately characterize the spatial dynamics of multi-energy interactions. Furthermore, TSTG adopts an end-to-end spatio-temporal joint optimization framework, achieving synchronous extraction and fusion of spatio-temporal features through an encoder–decoder architecture. Experimental results show that TSTG significantly outperforms existing methods in short-term load forecasting tasks, providing an effective solution for refined forecasting in integrated energy systems. Full article

Shallow foundations are widely used in both terrestrial and marine environments, supporting critical structures such as buildings, offshore wind turbines, subsea platforms, and infrastructure in coastal zones, including piers, seawalls, and coastal defense systems. Accurately determining the soil bearing capacity for shallow foundations presents a significant challenge, as it necessitates considerable resources in terms of materials and testing equipment, as well as a substantial amount of time to perform the necessary evaluations. Consequently, our research was designed to approximate the forecasting of soil bearing capacity for shallow foundations using machine learning algorithms. In our research, four ensemble machine learning algorithms were employed for the prediction process, benefiting from previous experimental tests. Those four models were AdaBoost, Extreme Gradient Boosting (XGBoost), Gradient Boosting Regression Trees (GBRTs), and Light Gradient Boosting Machine (LightGBM). To enhance the model’s efficacy and identify the optimal hyperparameters, grid search was conducted in conjunction with k-fold cross-validation for each model. The models were evaluated using the R² value, MAE, and RMSE. After evaluation, the R² values were between 0.817 and 0.849, where the GBRT model predicted more accurately than other models in training, testing, and combined datasets. Moreover, variable importance was analyzed to check which parameter is more important. Foundation width was the most important parameter affecting the shallow foundation bearing capacity. The findings obtained from the refined machine learning approach were compared with the well-known empirical and modern machine learning equations. In the end, the study designed a web application that helps geotechnical engineers from all over the world determine the ultimate bearing capacity of shallow foundations. Full article

The development of innovative and environmentally sustainable construction materials is a strategic priority in the context of the ecological transition and circular economy. Geopolymers and alkali-activated materials, derived from industrial and construction waste rich in aluminosilicates, are gaining increasing attention as low-carbon alternatives to ordinary Portland cement (OPC), which remains one of the main contributors to anthropogenic CO₂ emissions and landfill-bound construction waste. This review provides a comprehensive analysis of geopolymer-based solutions for building and architectural applications, with a particular focus on modular multilayer panels. Key aspects, such as chemical formulation, mechanical and thermal performance, durability, technological compatibility, and architectural flexibility, are critically examined. The discussion integrates considerations of disassemblability, reusability, and end-of-life scenarios, adopting a life cycle perspective to assess the circular potential of geopolymer building systems. Advanced fabrication strategies, including 3D printing and fibre reinforcement, are evaluated for their contribution to performance enhancement and material customisation. In parallel, the use of parametric modelling and digital tools such as building information modelling (BIM) coupled with life cycle assessment (LCA) enables holistic performance monitoring and optimisation throughout the design and construction process. The review also explores the emerging application of artificial intelligence (AI) and machine learning for predictive mix design and material property forecasting, identifying key trends and limitations in current research. Representative quantitative indicators demonstrate the performance and environmental potential of geopolymer systems: compressive strengths typically range from 30 to 80 MPa, with thermal conductivity values as low as 0.08–0.18 W/m·K for insulating panels. Life cycle assessments report 40–60% reductions in CO₂ emissions compared with OPC-based systems, underscoring their contribution to climate-neutral construction. Although significant progress has been made, challenges remain in terms of long-term durability, standardisation, data availability, and regulatory acceptance. Future perspectives are outlined, emphasising the need for interdisciplinary collaboration, digital integration, and performance-based codes to support the full deployment of geopolymer technologies in sustainable building and architecture. Full article

Accurately predicting stock prices is crucial for investment and risk management, but the non-stationarity of the financial market and the complex correlations among stocks pose challenges to traditional models (ARIMA, LSTM, XGBoost), resulting in difficulties in effectively capturing dynamic patterns and limited prediction accuracy. To this end, this paper proposes the Financial Spatio-Temporal Graph Attention Network (FSTGAT), with the following core innovations: temporal modelling through gated causal convolution to avoid future information leakage and capture long- and short-term fluctuations; enhanced spatial correlation learning by adopting the Dynamic Graph Attention Mechanism (GATv2) that incorporates industry information; designing the Multiple-Input-Multiple-Output (MIMO) architecture of industry grouping for the simultaneous learning of intra-group synergistic and inter-group influence; symmetrically fusing spatio-temporal modules to construct a hierarchical feature extraction framework. Experiments in the commercial banking and metals sectors of the New York Stock Exchange (NYSE) show that FSTGAT significantly outperforms the benchmark model, especially in high-volatility scenarios, where the prediction error is reduced by 45–69%, and can accurately capture price turning points. This study confirms the potential of graph neural networks to model the structure of financial interconnections, providing an effective tool for stock forecasting in non-stationary markets, and its forecasting accuracy and industry correlation capturing ability can support portfolio optimization, risk management improvement and supply chain decision guidance. Full article

Skid resistance, or friction, on a road surface is a critical parameter in functional highway assessments, given its direct relationships with safety and accident frequency. Therefore, road administrations must collect friction data across their road networks to ensure safe roads for users. In addition, having a predictive model of skid resistance for each road section is essential for an efficient pavement management system (PMS). Traditionally, road authorities disregard rural roads, since they are more focused on freeways and traffic-intense roads. This study develops a model for predicting minimum-available skid resistance, which occurs in summer, measured using the Sideway-force Coefficient Routine Investigation Machine (SCRIM), on bituminous pavements in the single-carriageway road network of the Province of Gipuzkoa, Spain. To this end, traffic volume data available in the PMS of the Provincial Council of Gipuzkoa, such as the annual average daily traffic (AADT) and the AADT of heavy vehicles (AADT.HV), were uniquely used to forecast skid-resistance values collected in summer. Additionally, a methodology for eliminating outliers is proposed. Despite the simplicity of the model, which does not include information about the materials at the surface layer, a coefficient of determination (R²) of 0.439 was achieved. This model can help road authorities identify the roads for which lower skid-resistance values are most likely to occur, allowing them to focus their attention and efforts on these roads, which are key infrastructure in rural areas. Full article

Transfers in professional football (soccer) are risky investments because of the large transfer fees and high risks involved. Although data-driven models can be used to improve transfer decisions, existing models focus on describing players’ historical progress, leaving their future performance unknown. Moreover, recent developments have called for the use of explainable models combined with methods for uncertainty quantification of predictions to improve applicability for practitioners. This paper assesses explainable machine learning models in a practitioner-oriented way for the prediction of the future development in quality and transfer value of professional football players. To this end, the methods for uncertainty quantification are studied through the literature. The predictive accuracy is studied by training the models to predict the quality and value of players one year ahead, equivalent to one season. This is carried out by training them on two data sets containing data-driven indicators describing the player quality and player value in historical settings. In this paper, the random forest model is found to be the most suitable model because it provides accurate predictions as well as an uncertainty quantification method that naturally arises from the bagging procedure of the random forest model. Additionally, this research shows that the development of player performance contains nonlinear patterns and interactions between variables, and that time series information can provide useful information for the modeling of player performance metrics. The resulting models can help football clubs make more informed, data-driven transfer decisions by forecasting player quality and transfer value. Full article

The recording of minerals extracted from a deposit is crucial for effective planning, exploitation management, and compliance with legal requirements. It also enables improved workplace safety and the minimization of negative environmental impact. Automation in mining optimizes exploitation, transportation, and data management processes, resulting in better forecasting, more accurate resource calculations, and reduced operational costs. The usage of geographic information system tools facilitates data modeling and analysis, enhancing monitoring and mining exploitation management. This paper presents the classical approach to determining the volume of extracted minerals and proposes GIS-based tools for the automation of the volume calculation process. The automation of the process is presented both from a theoretical perspective, providing requirements and parameters for individual calculation procedures, and from a practical perspective, using the example of a typical open pit mine, where the procedure is implemented starting from field measurements, carrying out calculations, and ending with visualization and interpretation. The study highlights the benefits of automating the calculation procedure for the volume of extracted minerals, including task execution acceleration, increased efficiency, reduced calculation time, and minimized human error. This ultimately leads to more precise and consistent results. Full article

China’s carbon peak and neutrality transition is critically constrained by the severe talent shortage and structural inefficiencies in higher education. This study systematically investigates the current status of “dual-carbon” talent cultivation and demand in China, leveraging annual “dual-carbon” talent cultivation data from universities nationwide. By applying the GM(1,1)-ARIMA hybrid forecasting model, it projects future national “dual-carbon” talent demand. Key findings reveal significant regional disparities in talent cultivation, with a pronounced mismatch between industrial demands and academic supply, particularly in interdisciplinary roles pivotal to decarbonization processes. Forecast results indicate an exponential growth in postgraduate talent demand, outpacing undergraduate demand, thereby underscoring the urgency of advancing high-end technological research and development. Through empirical analysis and innovative modeling, this study uncovers the structural contradictions between “dual-carbon” talent cultivation and market demands in China, providing critical decision-making insights to address the bottleneck of carbon-neutral talent development. Full article

This study investigates the capability of Gaussian process regression (GPR) models in the probabilistic forecasting of water flow and depth in a combined sewer system. Traditionally, deterministic methods have been implemented in sewer flow forecasting and anomaly detection, two crucial techniques for a good wastewater network and treatment plant management. However, with the uncertain nature of the factors impacting on sewer flow and depth, a probabilistic approach which takes uncertainties into account is preferred. This research introduces a novel use of GPR in sewer systems for real-time control and forecasting. To this end, a composite kernel is designed to capture flow and depth patterns in dry- and wet-weather periods by considering the underlying physical characteristics of the system. The multi-input, single-output GPR model is evaluated using root mean square error (RMSE), coverage, and differential entropy. The model demonstrates high predictive accuracy for both treatment plant inflow and manhole water levels across various training durations, with coverage values ranging from 87.5% to 99.4%. Finally, the model is used for anomaly detection by identifying deviations from expected ranges, enabling the estimation of surcharge and overflow probabilities under various conditions. Full article

Accurate prediction of pedestrian movements is vital for autonomous driving, smart transportation, and human–computer interactions. To effectively anticipate pedestrian behavior, it is crucial to consider the potential spatio-temporal interactions among individuals. Traditional modeling approaches often depend on absolute position encoding to discern the positional relationships between pedestrians. Unfortunately, this method overlooks relative spatio-temporal relationships and fails to simulate ongoing interactions adequately. To overcome this challenge, we present a relative spatio-temporal encoding (RSTE) strategy that proficiently captures and analyzes this essential information. Furthermore, we design a multi-spatio-temporal graph (MSTG) modeling technique aimed at modeling and characterizing spatio-temporal interaction data across several individuals over time and space, with the goal of representing the movement patterns of pedestrians accurately. Additionally, an attention-based MSTT model has been developed, which utilizes an end-to-end approach for learning the structure of the MSTG. The findings indicate that an understanding of an individual’s preceding trajectory is crucial for forecasting the subsequent movements of other individuals. Evaluations using two challenging datasets reveal that the MSTT model markedly outperforms traditional trajectory-based modeling methods in predictive performance. Full article

Background: this research presents an integrated approach to enhancing the sustainability of plastic bag supply chains in Indonesia by addressing critical issues related to ineffective post-consumer waste management and low recycling rates. The objective of this study is to develop a combined framework of material flow analysis (MFA) and sustainable supply chain planning to improve demand forecasting and inflow management across the plastic bag lifecycle. Method: the research adopts a quantitative method using the XGBoost algorithm for forecasting and is supported by a polymer-based MFA framework that maps material flows from production to end-of-life stages. Result: the findings indicate that while production processes achieve high efficiency with a yield of 89%, more than 60% of plastic bag waste remains unmanaged after use. Moreover, scenario analysis demonstrates that single interventions are insufficient to achieve circularity targets, whereas integrated strategies (e.g., reducing export volumes, enhancing waste collection, and improving recycling performance) are more effective in increasing recycling rates beyond 35%. Additionally, the study reveals that increasing domestic recycling capacity and minimizing dependency on exports can significantly reduce environmental leakage and strengthen local waste management systems. Conclusions: the study’s novelty lies in demonstrating how machine learning and material flow data can be synergized to inform circular supply chain decisions and regulatory planning. Full article