Performance Measurement Research Articles

Do transportation infrastructure contracts enhance spending efficiency? Examining state highway construction programs

This study examines the impact of state highway construction contracts on state spending efficiency controlling for production structure, service demands, and situational factors. The theoretical argument is that because highway construction projects are relatively large in scale, complex, and can be monitored through objective performance measurement, state highway construction programs may save government production costs through contracts. Contracting helps highway producers achieve efficiency by optimizing production size based on workload and task complexity. The unit of analysis is 48 state governments’ highway construction contracts from 1998 to 2008. Through a two-stage analysis method including a Total Function Productivity (TFP) index and system dynamic panel data analysis, the results suggest that highway construction contracts enhance state highway spending efficiency, especially for large-scale construction projects.

Traffic Flow Prediction Using Novel Spatial-Temporal Multi-Head Attention Graph Convolution Networks

For urban development, accurate traffic flow prediction is crucial. Traffic flow prediction data can be used for optimizing public transportation systems, reducing congestion, planning road networks, and improving overall city infrastructure management. Despite numerous spatiotemporal methods used to forecast traffic flow, information aggregation at each site neither produce optimal results, because of the use non-Euclidean distance data. For more accurate results of traffic flow prediction, we propose a spatial-temporal multi-head attention graph convolution network (STMAGCN). It extracts temporal features using temporal convolutional layers and enriches nodes with spatial attention and multi-head attention mechanisms. By combining them, we can extract space characteristics more effectively. We also introduce a novel graph structure construction method that employs graph-based transformations to convert non-connected data points into a comprehensive graph structure, facilitating the effective application of our STMAGCN. This approach improves the predictive efficiency and accuracy of our model by ensuring seamless integration of disparate data sources. Experimental results indicate that the preferential extraction of spatiotemporal data significantly enhances traffic flow prediction accuracy. Compared with the latest models, our performance on the California Department of Transportation Performance Measurement System (PeMS) PeMS04 and PeMS08 datasets shows a slight improvement, with an average increase of 16.8%. Additionally, the average improvement in time is approximately 445.95%.

University Management System

University Management System (UMS) is designed to automate the everyday operations of administrative, academic functions in higher degree education institutions. UMS are using integrated features like online registration for students, administration file maintenance of colleges faculty details course scheduling fee transactions academic performance monitoring in simple manner. We have seen an incredible change in thesesystems over the years, largely due to increased demands for digital transformation from schools. Today, there are cloud-based access to modern UMS solutions in place which allows institutions manage their data efficiently while retaining it on secure physical storage. These systems also come with mobile capabilities that are very useful in allowing students and faculty to remain up-to-date on the latest information. Advanced data analytics has become a critical ingredient as well, enabling administrators to balance the scales through academic planning, resource allocation and performance measurement. Education institutions are increasingly using digital ways of working and UMS have enabled these organizations to streamline processes, strengthen communication channels thus making the educational journey better for all stakeholders. Keywords— University Management System (UMS), higher education, administrative automation, academic processes, cloud-based access, mobile-friendly interface, data analytics, student registration.

Measurement of government performance in literature review

Government performance means the results of government work. Its use is to evaluate government accountability, decision-making, efficiency, effectiveness, transparency, and achievement of goals. Purpose: This paper aims to explore the understanding of performance measurement tools commonly used in government, the reasons for using them, and the implementation of performance measurement in Indonesia. Method: This study uses a meta-synthesis method, an integrative review approach from 2000–2021, in the Scopus database using the keywords measurement system, performance measurement, performance measurement government, measurement system government. Results and Discussion: The final sample consisted of 23 studies, and the results showed that the most commonly used performance measurement was the balanced scorecard. This is because the balanced scorecard is able to explain the vision, mission, strategy, results, and operational actions, so that it can achieve local government goals. Research implications: Insight into government performance measurement can be used to determine the strengths and weaknesses of various performance measurement tools so that the government can implement performance measurement tools that are more appropriate for its government. Originality/Value: This study offers an adaptation of existing methods to measure government performance more effectively. In addition, this study focuses on the context of developing countries, which can provide new contributions to the literature.

Abstract 4137264: Subclassification of Peripheral Artery Disease Based on Muscle Force: Leg Failure with Reduced Muscle Force and Leg Failure with Preserved Muscle Force

Introduction: All claudicating patients with peripheral artery disease (PAD) exhibit leg failure, as evidenced by distinct walking impairment. However, based on a detailed evaluation of leg biomechanics, some claudicating patients maintain preserved leg muscle strength, while others demonstrate reduced strength. In this study, we aim to compare the clinical and physiological characteristics of the two groups. Methods: Fifty-four claudicating patients were categorized into two groups: leg failure with preserved muscle force (LEF-PF) versus leg failure with reduced muscle force (LEF-RF), based on the peak torque (PT) from performing maximum isometric plantar flexion contractions on a Biodex system. All patients underwent history and physical exam, measurement of leg hemodynamics, walking performance and quality of life, and gastrocnemius biopsy to evaluate the histology and biochemistry of the affected leg muscles. We used independent samples T test and chi-square test to compare the two groups. Results: LEF-PF (N=19, PT 92±15 Nm) and LEF-RF (N=35, PT 49±13 Nm) had similar demographic and comorbidity profiles. While ankle-brachial index didn't differ between the two groups, during the reactive hyperemia test, ankle pressure of LEF-RF dropped more significantly (p=0.021) upon occlusion release and took longer to recover (p=0.023). Both groups performed similarly on the Gardner Treadmill test, but LEF-PF performed better in the 6-Minute Walking test (p=0.062). LEF-PF also trended towards higher scores on the modified falls efficacy scale (p=0.056). Histologically, gastrocnemius muscle from LEF-PF had larger myofibers, while biochemically, they exhibited lower mitochondrial respiratory activity dependent on complex-II compared to LEF-RF. Conclusion: Our study presents a novel approach to understanding the physiology of claudication, focusing on the strength of the ischemic muscles. Despite comparable demographics and hemodynamics at rest, notable differences separate the two groups during stress-induced hemodynamics, walking performance, myofiber histology, and mitochondrial biochemistry. Further investigations are necessary to elucidate the underlying mechanisms behind the differences in muscle strength among patients with claudication. Such insights may enhance our understanding of PAD physiology and potentially correlate with variations in mortality rates, progression to more advanced PAD stages, and responses to different therapeutic interventions.

Case study of a world hour record simulation in an elite cyclist: Insight into task failure.

The 'cycling hour-record' is one of the most prestigious events in cycling. However, little detailed analysis of such attempts is available. In preparation for a successful cycling hour-record attempt, an elite cyclist performed a full-hour simulation to provide insights into performance, physiological, aerodynamic and biomechanical limitations that could be identified in the preparation for a subsequent official attempt. Performance (speed, lap time, power and cadence), physiological (heart rate and estimated body temperature), aerodynamic (CDA, helmet angle, rotation and rock) and biomechanical (helmet, thigh and foot position changes) measurements were made throughout the attempt, in which an even-pacing strategy was employed where the point of task failure was defined as the lap which the rider could no longer perform at the targeted lap split (16.6s) or quicker. The cyclist did not achieve the target distance (54,000m) during the simulation. The final distance achieved for the hour was 53,250m. Task failure occurred at 38min and 33s (lap 139/34,750m) into the simulation. Notably, there was a decrease in power output, accompanied with an increase in the estimated body temperature, changes in pedalling kinematics and an increase in aerodynamic drag. The reduction in performance (leading to task failure) during a cycling hour record simulation is underpinned by a decrease in power output as well as an increase in aerodynamic drag due to biomechanical changes in the cycling technique.

Leveraging Industry 4.0 for marketing strategies in the medical device industry of emerging economies.

This research investigates the impact of the technologies through an integrated Fuzzy Analytic Hierarchy Process (F-AHP) and VIKOR model to determine their viability as a marketing strategy for business growth in the Prosthetics and Orthotics (P&O) medical device industry. Based on expert data, the study evaluates the integration of Industry 4.0 technologies and ranks them by priority for marketing performance: "Artificial Intelligence & Machine Learning" (0.3295), "Internet of Things" (0.4325), "Virtual Reality & Augmented Reality" (0.5000), and "Big Data" (0.5704). The findings confirm that the proposed model effectively Leverage the industry 4.0 technologies to enhance marketing strategies, using Pakistan as a case study. This research serves as a blueprint for addressing the integration of next-generation marketing tools, technological advancements, and business growth challenges in the P&O industry. It contributes to the growing body of knowledge on the role of technology in improving healthcare marketing and business performance in emerging markets. Future research should explore the broader impact of Industry 4.0 technologies on healthcare marketing, focusing on strategic decision-making, efficient marketing resource allocation, and performance measurement to ensure sustained success in the P&O medical device sector. The study concludes that prioritizing Industry 4.0 technologies, however, limitations such as sample size, data collection constraints, and model complexity may Leverage the generalizability of the results. Future research should address these limitations by expanding the scope of expert involvement and exploring the application of this model in other emerging economies, to further validate the findings and assess the broader implications for healthcare marketing strategies globally.

Evidence Provided by the Literature on the Use of Crossfit Elements in Sport

In recent years, scientific literature has increasingly examined the use of CrossFit in various sports. This paper aims to analyze this literature through bibliometric methods. Searches in the Scopus database from 2000 to January 30, 2024, yielded 567 relevant results. The analysis identified publication dates, prominent authors, affiliations, and key publications, aligning findings with UN goals. Around 75% of articles were published in the last 5 years, indicating rising interest. Results were categorized into sports branches, performance measurement, injuries, nutrition, motivation, and benefits.

Spatio-temporal attention based collaborative local–global learning for traffic flow prediction

Traffic flow prediction is crucial for intelligent transportation systems (ITS), providing valuable insights for traffic control, route planning, and operation management. Existing work often separately models the spatial and temporal dependencies and primarily relies on predefined graphs to represent spatio-temporal dependencies, neglecting the traffic dynamics caused by unexpected events and the global relationships among road segments. Unlike previous models that primarily focus on local feature extraction, we propose a novel collaborative local–global learning model (LOGO) that employs spatio-temporal attention (STA) and graph convolutional networks (GCN). Specifically, LOGO simultaneously extracts hidden traffic features from both local and global perspectives. In local feature extraction, a novel STA is devised to directly attend to spatio-temporal coupling interdependencies instead of separately modeling temporal and spatial dependencies, and to capture in-depth spatio-temporal traffic context with an adaptive graph focusing on the dynamics in traffic flow. In global feature extraction, a global correlation matrix is constructed and GCNs are utilized to propagate messages on the obtained matrix to achieve interactions between both adjacent and similar road segments. Finally, the obtained local and global features are concatenated and fed into a gated aggregation to forecast future traffic flow. Extensive experiments on four real-world traffic datasets sourced from the Caltrans Performance Measurement System (PEMS03, PEMS04, PEMS07, and PEMS08) demonstrate the effectiveness of our proposed model. LOGO achieves the best performance over 18 state-of-the-art baselines and the best prediction performance with the highest improvement of 6.06% on the PEMS07 dataset. Additionally, two real-world case studies further substantiate the robustness and interpretability of LOGO.

Higher Percentage of Virtual Primary Care Associated With Minimal Differences in Achievement of Quality Metrics.

To test the impact of virtual care usage on quality metrics used for performance measurement. Virtual care improves access to primary care; however, the quality of care must not be adversely impacted by its use. This is a mixed-design etiologic study using data from patients receiving primary care in a large, regional health system from January 2020 through December 2021. Eligible patients had at least one primary care contact. Eligible physicians had 10 or more patient contacts. The quartile of virtual visits per physician per month is calculated as the percentage of total visits conducted by phone or video (Q1 is the lowest). Six metrics used for value-based reimbursement were chosen for modeling with generalized linear mixed models. The data included 200,090 patients of 683 physicians in 42 clinics over 24 months. Virtual care usage peaked in April 2020 at 78% and then stabilized at 18%. The blood pressure metric was met in 66% (95% CI: 63%-69%) of physician months in Q1 and 65% (95% CI: 63%-68%) in Q4 (P = 0.003). The hemoglobin A1c metric was met in 73% (95% CI: 70%-76%) of physician months in Q1 and 72% (95% CI: 69%-75%) in Q4, not a significant difference. Breast cancer screening completion and colon cancer screening completion did not differ across virtual care quartiles. Medicare annual wellness visits were completed in 55% (95% CI: 50%-60%) of Q1 physician months and 54% in each of Q2, Q3, and Q4 (P < 0.0001). Some quality metrics were modestly impacted by high virtual primary care usage; the absolute differences in rates were small. This may provide reassurance to physicians and their health systems that telemedicine use may not adversely impact quality metrics.

Multi-objective time-cost-safety risk trade-off optimization for the construction scheduling problem

PurposeThe purpose of this study is to enable the planning of construction projects with simultaneous consideration of time, cost and safety risks. It also aims to improve the decision-making process by evaluating the effectiveness of the Rao-2 algorithm in solving multi-objective time-cost-safety risk problems. In the end, this model is designed to support project managers in enhancing management approaches by addressing project challenges and constraints more efficiently.Design/methodology/approachIn this study, the Rao-2 algorithm, along with Grey Wolf Optimization (GWO) and Whale Optimization algorithm (WOA), were improved using the crowding distance-based non-dominated sorting method. Rao-2 was first compared to GWO and WOA. Subsequently, it was compared with well-established algorithms in the literature, including genetic algorithm (GA), particle swarm optimization (PSO) and differential evolution (DE). The C-metric, hypervolume and spread metrics were employed for performance measurement. The performance of the algorithms was evaluated on four case studies consisting of 11, 13, 18 and 25 activities.FindingsThe results revealed that Rao-2 performs better than other algorithms as the number of activities increases, when compared using the Hypervolume, Spread and C-metric measures. In terms of performance measures, the GWO algorithm outperformed Rao-2 in some evaluation metrics for the instance involving 11 activities. However, as the number of activities grew, the Rao-2 method consistently generated higher-quality Pareto fronts and outperformed GWO and WOA in all evaluation metrics. The solutions generated by Rao-2 were also superior to those obtained from GA, PSO and DE in all case studies, further demonstrating the capability of our framework to produce a wide range of optimal solutions with high diversity across different case studies.Originality/valueThis research demonstrates that Rao-2 not only improves solution quality when generating Pareto fronts but also achieves better results with fewer function evaluations compared to GA, PSO and DE. The algorithm's efficiency makes it particularly well-suited for optimizing time, cost and safety risks in large-scale construction projects, which in turn positions Rao-2 as a better choice for such projects by producing superior results compared to other algorithms. By providing high-quality solutions with reduced computational demands, Rao-2 offers a faster and more resource-efficient tool for decision-making, contributing to advancements in both the theory and practice of construction project management.

Research on the design and application of Analog-to-Digital converters

Abstract. In the modern digital era, the conversion of analog signals to digital data is a crucial process for a wide range of applications in industries such as consumer electronics, medical devices, and telecommunications. The accelerated evolution of digital technology is driving a surge in demand for high-performance analog-to-digital converters (ADCs). The paper presents a comprehensive analysis of analog-to-digital converters, emphasizing their types, applications, performance measurements, and most recent developments. In the paper, emphasis is placed on the mechanisms of different ADC architectures, such as Successive Approximation Register (SAR), Sigma-Delta (-), and Flash ADCs, and their respective strengths and weaknesses are evaluated. In terms of technique, it synthesizes and summarizes research results from multiple academic sources and technical papers. In addition to underscoring the pivotal concerns of ADC design, including power consumption, resolution, and sample rate, the research delves into novel approaches that leverage the use of advanced materials and techniques. The research results demonstrate the continued development of ADC technology and its critical function in improving the effectiveness and performance of digital systems.

Portfolio Assessment and Contact-Hour-Based Evaluation (ChbE) in Architectural Education

In Architectural Education, the assessment of design courses involves the collective participation of the studio masters, the jurors, and the department board to assess the student’s ability to interpret and proffer solutions from thought processes to design proposals. Unfortunately, the assessment process is heterogeneous in the National Board for Technical Education (NBTE) accredited Schools of Architectural Technology, as institutions deploy atypical assessment scoresheets with varying variables and concomitant weightings. In most cases, such non-standard assessment variables laud students’ cognitive ability to the detriment of their affective and psychomotor persuasions in design development. This has structurally flawed the measurement of performance in the design ethos of Architectural Technology students. This issue necessitated the inception of this study towards defining the limits of assessment and evaluation protocols of design courses in Architectural Education and determining the variables that constitute criteria for assessment of portfolios. The study involved four (4) NBTE-accredited institutions that offer Architectural Technology. The methodology adopted structured questionnaires and interviews, observations, comparative analysis of scoresheets from such institutions, and archival retrieval of policy documents. Purposive sampling was adopted to administer structured questionnaires to 45 lecturers out of 63 lecturers serving in these institutions thus constituting 71.43% of the academic staff population. Results revealed poor evaluation of design miscellany in Architectural Education and the adoption of an incompatible assessment structure of design courses in the October 2020 NBTE curriculum for Architectural Technology. The study recommended the adoption of a standard Portfolio Assessment Sheet (PAS) and nouveau Contact-hour-based Evaluation module (ChbE) for measuring competence in design miscellany. A review of the assessment structure of design courses in the October 2020 NBTE curriculum and appraisal of the subsisting policy on Architectural Education were also considered imperative.

Enhancing and Explaining Energy Management in Smart Manufacturing Production via ML Within the Context of Industry 4.0

Objective: The aim of this research is to identify the main barriers faced by managers and practitioners in the implementation of KM initiatives in the context of Chilean organisations and to describe the ways in which these problems are addressed by them. Theoretical Framework: This research addresses the organisational, political, social and cultural barriers faced by KM initiatives in developing country contexts. Method: This study was based on the review of twenty-six cases of knowledge management initiatives implemented between 2010 and 2022 in Chile, which were described and analysed using qualitative methodology Results and Discussion: It was found that the most important barriers to carrying out these initiatives coincide, to a large extent, with some of the characteristics of Chilean culture and society described in the literature of organisational studies, such as: lack of trust; large power distance in the organisational structure; lack of communication skills; paternalistic organisational culture, to mention a few. Also, one aspect worth highlighting is the traditional business culture that tends to focus on a "low cost" strategy and exploitation of raw materials, rather than investing in innovation. Scarce resources, problematic performance measurement and lack of knowledge management training for employees and managers were also mentioned as adverse factors. Research Implications: The relevance of this study lies in the fact that it provides insight into the main barriers faced by knowledge management initiatives in developing countries, specifically in a region such as Latin America, thus contributing to the development of this field in this region. Originality/Value: It is known that many of the new initiatives in the field of KM do not thrive or face difficult start-up challenges, limiting the number of successful initiatives in this country. This study contributes to a better understanding of the problems faced by managers carrying out KM initiatives, thus enabling them to develop better strategies and solutions for the implementation of KM projects.

Association between weight and body composition changes with falls risk in the Malaysian Elders Longitudinal Research (MELoR) study

ObjectiveBoth changes in body composition and increased fall risk occur with increasing age. While weight management may be considered a component of falls prevention, the long-term consequences of changes in...

Different sustainability endgames: Institutional logics in the performance management of local governments

IMPACT Local governments operate in institutionally complex environments where overlapping values, rules and priorities shape their actions. This article explores how such complexity directly impacts the production of sustainability performance information. By examining various institutional logics within local governments, the study highlights how these differing perspectives lead to diverse approaches in managing sustainability performance. The article introduces three potential ‘endgames’—sustainability as a brand, strategy or raison d’être—each representing a distinct path for integrating sustainable development goals (SDGs) into local governance. These endgames provide local government managers, specialists, and sustainability co-ordinators, with a framework to reflect on the priorities and rationales behind current sustainability performance management practices. This article is essential reading for those seeking to develop meaningful performance measurement for the SDGs.

Measuring Supply Chain Performance Using the SCOR Model: A Case Study of CV. XYZ

This research discusses the design of a performance measurement system for supply chain management using the Supply Chain Operations Reference (SCOR) model developed by the Supply Chain Council. The SCOR model focuses on five main activities within a company's supply chain, namely plan, source, make, deliver, and return. All activities are linked with performance attributes such as reliability, responsiveness, flexibility, cost, and assets. The implementation of this model is applied to CV. XYZ, a company that operates in the production of plain and printed T-shirts, which is located in East Java. The research was conducted from January 2024 to September 2024. The results of the study identified that there are 35 performance indicators representing supply chain activities, with the following breakdown: 9 indicators for the plan activity, 7 indicators for the source activity, 6 indicators for the make activity, 7 indicators for the deliver activity, and 6 indicators for the return activity. The Analytical Hierarchy Process (AHP) method was used to weight each activity in the SCOR model. Based on the results, the highest priority was the source activity (0.252), followed by deliver (0.222), plan (0.221), make (0.209), and return (0.095). From the scoring system, it was found that 23 indicators have a green score, 8 indicators have a yellow score, and 4 indicators have a red score. The limitation of this research is that one SCOR activity, "enable," was not analyzed. Therefore, it is recommended that future research analyze all SCOR activities.

Designing and testing a sustainability soft regulation system: rating and ranking the wild animal welfare practices of tourism distribution channels

ABSTRACT The lack of standards pertaining to wild animal welfare practices in tourism creates an opportunity to introduce soft regulations to improve public accountability and foster positive change. We developed a rating and ranking system to embed the ideologies of a non-governmental organisation to improve wild animal welfare practices, acting as institutional entrepreneurs to gain pragmatic legitimacy within tourism distribution channels. The new system aims to create change in three ways: (i) informing, (ii) ordering, (iii) surveyance and control of wild animal welfare. We candidly report on the methodological choices behind the system’s design and implementation. We evaluate tourism distribution channels on four aspects, their: (i) policy and commitment statements, (ii) target setting and performance measurement, (iii) industry engagement and (iv) consumer awareness initiatives. An analysis of data from 12 international companies in 2020 and 2022 demonstrates how the system allows for shallow engagement in wild animal welfare practices, yet also offers the potential for significant improvements. We reflect on three methodological challenges: (i) quantification and commensuration during system design, (ii) the availability and credibility of data sources and (iii) the transparency of results.

Heterogeneous Online Computational Platform for GEM-Based Plasma Impurity Monitoring Systems

The fusion energy research field presents many intricate challenges that require resolution. Many diagnostic systems employed in experiments are approaching the limits of current technology. Implementing efficient measurements requires using an appropriate set of tools to facilitate the optimal utilization of hardware. Fusion energy measurements must provide low latency processing with the capacity for future improvements and the ability to handle complex data flows efficiently. The presented work addresses these requirements and describes the implementation of a high-performance, low-latency software platform with convenient development for soft X-ray (SXR) plasma impurities emission tracing—the Asynchronous Complex Computation Platform (AC2P). This article presents the architectural design, implementation details, and performance and latency measurements based on the raw data acquired from the WEST tokamak and laboratory tests. AC2P provides the tools to develop low-latency, multi-core, multi-device complex data flow graph scale-up solutions for measuring impurities in hot plasmas. The system has been designed to operate online during experiments, calculate the energy distribution, position and occurrence time of SXR photons, monitor the data stream’s quality and archive any abnormalities for subsequent offline verification and algorithm improvement. This article presents AC2P, which operates as part of the SXR measurement system on the WEST tokamak.

A Comprehensive Approach to Fault Classification of Helicopter Engines with Adaboost Ensemble Model

This work is based on the PHM North America 2024 Conference Data Challenge’s datasets of Helicopter turbine engine performance measurements. These datasets were large and moderately imbalanced. For dealing with these challenges, we demonstrate significant tools covering feature engineering, augmentation and selection, model exploration, visualizations, model explainability and confidence margin estimation. This work was performed in its entirety using MATLAB. All these tools will be generally applicable to data-driven modeling and prediction of health to real life applications. Initially, we explored the 742k observations in the training set, noting a 60-40 split between healthy and faulty labels, and identified two major operational clusters within the data. We enhanced the dataset by removing duplicates and engineered new features based on domain knowledge, expanding the feature set to 242 dimensions. However for the torque margin estimation, we trained a regression model on a limited subset (18 features), which includes engineered features using domain knowledge, quadratic terms and linear interaction between all the terms. For the final submission, we utilized a stepwise linear regression model to optimize feature selection. This approach achieved a perfect regression score on test data, validated by a consistent torque margin residual range of +/- 0.5%. The model's RMSE and MAE metrics were optimal for employing a normal distribution probability density function. For , we reduced the feature set to 58 using dimensionality reduction techniques and balanced the data with upsampling and down-weighing the minority class. We employed ASHA (Asynchronous Successive Halving Algorithm) in conjunction with AutoML to efficiently determine the most suitable model family, significantly saving compute time. Subsequently, we trained ensemble models, including bagged tree and AdaBoost (Adaptive Boosting), which minimized false negatives and positives, achieving robust classification performance. This was particularly critical given the high penalty for false negatives in the data challenge. The MathWorks team score on Testing Data was 0.9686 at the close of competition. This was further improved to 0.9867. Our approach demonstrates the effectiveness of combining strategic data processing, feature engineering, and model selection to enhance predictive accuracy in complex operational datasets.