Performance Model Research Articles

Capsular attention Conv-LSTM network (CACN): A deep learning structure for crop yield estimation based on multispectral imagery

Precise prediction of agricultural production output is crucial for farmers, policymakers, and the Farming-related industry. This article introduces a novel methodology to crop yield forecasting using a capsular neural network equipped with Conv-LSTM and attention mechanism. Our model combines the strengths of 3DCNN, and Conv-LSTM, which can capture the temporal dependencies and 3D features of crop yield data, and attention mechanism, which Can prioritize the most significant characteristics for making predictions. We evaluated CACN on a sizable collection of data of soybean crop yield in the United States from 2003 to 2019 and evaluated against various cutting-edge deep learning models. The outcomes indicate that our suggested approach surpasses other models in performance in terms of RMSE, correlation coefficient, and prediction error map. Specifically, our model achieved approximately 14 % improvement in terms of RMSE, compared to the state-of-the-art model Deep-Yield. Our model also demonstrated the ability to extract more meaningful features and capture the complex relationships between crop yield data and meteorological variables. Overall, our proposed method shows great potential for accurate and efficient crop yield forecasting and can be applied to other crops and regions.

Modeling and Prediction of Surface Roughness in Ball End Milling of Oxygen-Free High Conductivity Copper Using Adaptive Neuro Fuzzy Inference System

Oxygen Free High Conducting Copper (OFHC) is one of the reasons materials for numerous applications due to its high thermal conductivity, great machinability, and great quality In this paper, an adaptive neuro-fuzzy inference system (ANFIS) is developed as a predicting model of the machining performance of OFHC measured for surface roughness in terms of process parameters, namely, the cutting feed rate or feed per tooth, axial depth of cut, radial depth of cut, and the cutting speed.

Banks’ financial performance analysis: an experience from an Islamic economy

Purpose This study aims to analyze the performance efficiency of private commercial banks in Iran as an Islamic economy. Design/methodology/approach A meticulously chosen sample comprising 15 private banks from 2020 to 2023 has been examined. This study uses data envelopment analysis (DEA) models, specifically the Constant Returns to Scale (CCR) with input orientation and the Variable Returns to Scale (BCC) with input orientation, to scrutinize performance efficiency relative to the banking sector’s average efficiency ratio. Findings The findings indicate that the performance of decision making units is superior in BCC models when determined with CCR models. The CCR-I analysis spanning 2020–2023 reveals that only two banks consistently demonstrated full efficiency performance, attaining a 100% efficiency score across all years. The observed fluctuations in banks’ efficiency performance are attributed to disparities between the growth or reduction in inputs and the corresponding augmentation or diminution in outputs. Research limitations/implications This paper has some limitations. First, selecting inputs and outputs for the DEA models is crucial, as this study’s results are sensitive to these choices. Second, this research focused exclusively on the efficiency of private banks and did not assess the efficiency of governmental banks. Practical implications The insights from this study can be valuable for bank managers in evaluating performance, regulators in maintaining financial stability and investors and market analysts in enhancing their investment strategies through bank performance analysis. Originality/value To the best of the authors’ knowledge, this study is the first to propose a framework for evaluating and forecasting commercial bank performance in an Islamic background. This model for prediction performance would be helpful for future research on the banking sector. In addition to improving Islamic banking literature, this research contributes to the literature regarding the analysis methods usually used.

High-Resolution SAR-to-Multispectral Image Translation Based on S2MS-GAN

Synthetic aperture radar (SAR) has been extensively applied in remote sensing applications. Nevertheless, it is a challenge to process and interpret SAR images. The key to interpreting SAR images lies in transforming them into other forms of remote sensing images to extract valuable hidden remote sensing information. Currently, the conversion of SAR images to optical images produces low-quality results and incomplete spectral information. To address these problems, an end-to-end network model, S2MS-GAN, is proposed for converting SAR images into multispectral images. In this process, to tackle the issues of noise and image generation quality, a TV-BM3D module is introduced into the generator model. Through TV regularization, block-matching, and 3D filtering, these two modules can preserve the edges and reduce the speckle noise in SAR images. In addition, spectral attention is added to improve the spectral features of the generated MS images. Furthermore, we construct a very high-resolution SAR-to-MS image dataset, S2MS-HR, with a spatial resolution of 0.3 m, which is currently the most comprehensive dataset available for high-resolution SAR-to-MS image interpretation. Finally, a series of experiments are conducted on the relevant dataset. Both quantitative and qualitative evaluations demonstrate that our method outperforms several state-of-the-art models in translation performance. The solution effectively facilitates high-quality transitions of SAR images across different types.

A comprehensive comparison of deep learning-based compound-target interaction prediction models to unveil guiding design principles

The evaluation of compound-target interactions (CTIs) is at the heart of drug discovery efforts. Given the substantial time and monetary costs of classical experimental screening, significant efforts have been dedicated to develop deep learning-based models that can accurately predict CTIs. A comprehensive comparison of these models on a large, curated CTI dataset is, however, still lacking. Here, we perform an in-depth comparison of 12 state-of-the-art deep learning architectures that use different protein and compound representations. The models were selected for their reported performance and architectures. To reliably compare model performance, we curated over 300 thousand binding and non-binding CTIs and established several gold-standard datasets of varying size and information. Based on our findings, DeepConv-DTI consistently outperforms other models in CTI prediction performance across the majority of datasets. It achieves an MCC of 0.6 or higher for most of the datasets and is one of the fastest models in training and inference. These results indicate that utilizing convolutional-based windows as in DeepConv-DTI to traverse trainable embeddings is a highly effective approach for capturing informative protein features. We also observed that physicochemical embeddings of targets increased model performance. We therefore modified DeepConv-DTI to include normalized physicochemical properties, which resulted in the overall best performing model Phys-DeepConv-DTI. This work highlights how the systematic evaluation of input features of compounds and targets, as well as their corresponding neural network architectures, can serve as a roadmap for the future development of improved CTI models.Scientific contributionThis work features comprehensive CTI datasets to allow for the objective comparison and benchmarking of CTI prediction algorithms. Based on this dataset, we gained insights into which embeddings of compounds and targets and which deep learning-based algorithms perform best, providing a blueprint for the future development of CTI algorithms. Using the insights gained from this screen, we provide a novel CTI algorithm with state-of-the-art performance.

Predicting hospital readmission in multimorbid patients with the use of AI: A systematic review

Abstract Background Multimorbid patients are at higher risk of hospital readmission due to the complex nature of their conditions. Identifying those who may be at particularly high risk would allow us to intervene early and potentially delay or prevent such readmissions occurring, thus reducing healthcare costs. We conducted a systematic review investigating the use of machine learning models in predicting 30-days unplanned hospital readmission of multimorbid patients. Methods This review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Guidelines, and was registered with PROSPERO (CRD42022373937). We searched MEDLINE, Embase, Web of Science and Cumulative Index to Nursing and Allied Health Literature (CINAHL). Included studies developed an AI model for prediction of 30-days hospital readmission for adult patients with two or more health conditions. The CHARM and PROBAST checklists for data extraction and bias assessment were used. The quality of included studies was assessed with the CASP checklist. Results Eighteen papers were eligible for inclusion. A total of 669 predictors were reported with an average of 37 used per model. Predictors were classified as modifiable and non-modifiable with the most common modifiable predictors being hospital length of stay, hypertension, anaemia and obesity. Average sensitivity and specificity of the models was reported to be 72% in 13 studies and 79.2 in 11 studies, respectively. Area under the curve (AUC) was reported in 13 studies, five of which were considered to have good discrimination power (AUC>0.8). Conclusions Machine learning models are capable of accurately predicting 30-days hospital readmission of multimorbid patients. Identifying modifiable predictors with highest weight allows for better planning and resource allocation to potentially reduce the risk of 30-days readmissions. An important area for future work would be the implementation of these high performance models in practice. Key messages • Hospital readmissions are preventable and identifying those at higher risk allows us to intervene early. • Machine learning models are capable of predicting 30-days readmission of multimorbid patients.

Late acquisition of erect hindlimb posture and function in the forerunners of therian mammals.

The evolutionary transition from early synapsids to therian mammals involved profound reorganization in locomotor anatomy and function, centered around a shift from "sprawled" to "erect" limb postures. When and how this functional shift was accomplished has remained difficult to decipher from the fossil record alone. Through biomechanical modeling of hindlimb force-generating performance in eight exemplar fossil synapsids, we demonstrate that the erect locomotor regime typifying modern therians did not evolve until just before crown Theria. Modeling also identifies a transient phase of increased performance in therapsids and early cynodonts, before crown mammals. Further, quantifying the global actions of major hip muscle groups indicates a protracted juxtaposition of functional redeployment and conservatism, highlighting the intricate interplay between anatomical reorganization and function across postural transitions. We infer a complex history of synapsid locomotor evolution and suggest that major evolutionary transitions between contrasting locomotor behaviors may follow highly nonlinear trajectories.

Design Considerations for Power-Efficient Fully Integrated 3:1 Switched Capacitor DC-DC Converter for PV Modules

This article presents a power-efficient DC-DC converter based on a switched-capacitor (SC) cell in power management systems supplied for fully integrated photovoltaic (PV) modules. These modules shall provide high-performance point-of-load voltage regulation. The primary objective of this study is to better utilize capacitance and switches by selecting a proper SC topology in order to improve the power efficiency of SC converters. A general steady-state performance model is investigated to optimize and compare a variety of SC DC-DC topologies. The investigation method relies on a charge-multiplier approach and considers the impact of area constraint on capacitors. To identify the most suitable topology for a given conversion ratio, the performance-limit metrics of SC converters are calculated. The analysis provides framework to determine optimum switch size and switching frequency for a two-phase 3:1 series–parallel converter for a target load current of 10 mA implemented on a 22 nm process technology. The results shows that a minimum of 250 MHz switching frequency is desirable for achieving a target efficiency greater than 85% while maintaining the minimum output voltage of 0.34 V. The analysis results are verified through MATLAB and PSpice-based simulations.

Comparative Study and Modeling of AlGaN/GaN Heterostructure HEMT and MOSHEMT Biosensors

This paper presents a comprehensive comparative study and analytical modelling of electrical performance of AlGaN/GaN high-electron-mobility transistor (HEMT) and metal-oxide-semiconductor high-electron-mobility transistor (MOSHEMT) biosensors. Sensing parameters such as the I-V characteristics and sensitivity parameter for biomolecules detection in the cavity region are taken into consideration. In this paper, the permittivity is varied according to the biomolecule to be sensed by the biosensor. The maximal variation of the electrical performance of the biosensor obtained is higher in HEMT as compared with MOSHEMT. The simulation results of the analytical model obtained by using MATLAB verified by a comparison with experimental data and atlas-technology computer aided design (Atlas-TCAD), and shown good agreement with each other. Thereby, we could improve the validity of the proposed model. The AlGaN/GaN HEMT have shown good sensing of 141.73 at biomolecular permittivity of 2.5 which can be used for biosensing applications effectively.

New hybrid system of PV/T, solar collectors, PEM electrolyzer, and HDH for hydrogen and freshwater production: Seasonal performance investigation

A new hybrid energy system of PV/T, flat plate collector, PEM electrolyzer, and humidification-dehumidification desalination is introduced here. A comprehensive mathematical model for the system seasonal performance is constructed and solved numerically using MATLAB software. A year-round comparison between the traditional PV/T performance and its concentrated counterpart indicates that the latter performs better compared to power, hydrogen, and freshwater production and system overall efficiency. The maximum monthly achieved PV/T efficiency, electrolyzer efficiency, desalination system Gain Output Ratio, and overall system efficiency are 52 %, 62 %, 4.8 %, and 44 %, respectively at a concentration ratio of 4 and maximum allowable PV temperature. The system produces 82,200 kWh of energy annually and provides 77,700 kWh and 1500 kWh to the electrolyzer and the electric heater, respectively. Moreover, the system produces an annual freshwater production of about 52,700 kg, out of which approximately 10,000 kg are consumed in the electrolysis process, and an annual hydrogen production of approximately 1120 kg.

Modeling Civil Aviation Emissions with Actual Flight Trajectories and Enhanced Aircraft Performance Model

Aviation emissions are continuously increasing along with the rapid development of air transportation, and results in the deterioration in regional air quality and the global climate. Accurate emission estimation is of great importance for relevant policies promotion and the sustainable development of the environment. Previous studies focused on the total emissions of a flight and lacked high precision in both spatial and temporal resolutions, especially aviation activities near ground. In this research, we propose an open-sourced emission calculation framework based on actual flight trajectories (TrajEmission), which calculates both the ground and airborne emissions simultaneously according to the configuration parameters, trajectory characteristics, and ambient conditions. We compare the emission results with five emission inventory methods. The results indicate that pollutant (nitrogen oxides, carbon monoxide, and unburned hydrocarbons) emissions in the landing and takeoff (LTO) cycle might usually be underestimated due to a lack of trajectory-based methods. In addition, in the overall results, the method based on the great circle route leads to an overestimation of 56.8% of pollutant emissions compared to the method based on actual routes. We also investigate the extent to which other factors could influence the emission results. To summarize, the TrajEmission framework can build inventories for the whole process of flight movements with high spatial–temporal resolutions and provide solid data support for environmental science and other related fields.

Macroscopic State-Level Analysis of Pavement Roughness Using Time–Space Econometric Modeling Methods

This paper used pavement condition data collected by the Federal Highway Administration (FHWA) between 2001 and 2006 aggregated by U.S. states to identify macroscopic factors affecting pavement roughness in time and space. To account for prior pavement conditions and preservation expenditure over time, time autocorrelation parameters were introduced in a spatial modeling scheme that accounted for spatial autocorrelation and heterogeneity. The proposed framework accommodates data aggregation in network-level pavement deterioration models. Because pavement roughness across different roadway classes is anticipated to be affected by different explanatory parameters, separate time–space models are estimated for nine roadway classes (rural interstate roads, rural collectors, urban minor arterials, urban principal arterials, and other freeways). The best model specifications revealed that different time–space models were appropriate for pavement performance modeling across the different roadway classes. Factors that were found to affect state-level pavement roughness in time and space included preservation expenditure, predominant soil type, and predominant climatic conditions. The results have the potential to assist governmental agencies in planning effectively for pavement preservation programs at a macroscopic level.

Sequential lexicon enhanced bidirectional encoder representations from transformers: Chinese named entity recognition using sequential lexicon enhanced BERT

Lexicon Enhanced Bidirectional Encoder Representations from Transformers (LEBERT) has achieved great success in Chinese Named Entity Recognition (NER). LEBERT performs lexical enhancement with a Lexicon Adapter layer, which facilitates deep lexicon knowledge fusion at the lower layers of BERT. However, this method is likely to introduce noise words and does not consider the possible conflicts between words when fusing lexicon information. To address this issue, we advocate for a novel lexical enhancement method, Sequential Lexicon Enhanced BERT (SLEBERT) for the Chinese NER, which builds sequential lexicon to reduce noise words and resolve the problem of lexical conflict. Compared with LEBERT, it leverages the position encoding of sequential lexicon and adaptive attention mechanism of sequential lexicon to enhance the lexicon feature. Experiments on the four available datasets identified that SLEBERT outperforms other lexical enhancement models in performance and efficiency.

Performance degradation modeling of civil aviation engines based on component characteristic map optimization

In order to provide a theoretical basis for the degradation of the air path performance of civil aviation engines at the unit level, the CFM56-3 engine was used as the research object. First, on the basis of using the characteristic map scaling method to obtain the component characteristic equation, the selection process of the scaling reference point of the fan general characteristic map was optimized, and a surface fitting method of the characteristic map was proposed to construct an engine component-level benchmark performance model that meets specific speed conditions under steady-state conditions. Then, by introducing the fault factor to generate the fault coefficient matrix, the deviation of the engine monitoring parameters as the component efficiency decreases was calculated, and compared with the fingerprint map data of the GE training manual, it was verified that the engine steady-state performance model that integrates the fan characteristic map scaling reference point optimization and the characteristic map surface fitting method has good accuracy and use in the analysis of air path performance degradation.

Explainable machine learning models for predicting the ultimate bending capacity of slotted perforated cold-formed steel beams under distortional buckling

This study develops explainable machine learning (ML) models to predict the ultimate bending capacity of cold-formed steel (CFS) beams with staggered slotted perforations, focusing on distortional buckling behavior. Utilizing a dataset from 432 non-linear finite element analysis simulations of CFS Lipped channels, ten ML algorithms, including four basic and six ensemble models, were evaluated. Ensemble models, specifically CatBoost and XGBoost, demonstrated superior accuracy, with test-set performances reaching a coefficient of determination (R2) of 99.9%, outperforming traditional analytical methods such as the Direct Strength Method (DSM). SHapley Additive Explanations (SHAP) were applied to highlight how features like plate thickness and root radius critically influence predictions. The findings underscore the enhanced predictive capabilities of ML models for structural performance, suggesting a significant potential to refine traditional design methodologies and optimize CFS beam designs.

An AutoEncoder enhanced light gradient boosting machine method for credit card fraud detection

Online financial transactions bring convenience to people’s lives, but also present vulnerabilities for criminals to embezzle users’ accounts and trick users into credit card fraud. Although machine learning methods have been adopted to detect anomalous transactions, it’s hard for a single machine learning method to achieve satisfying results with the increasing scale and dimensionality of financial datasets. In addition, for anomaly detection of financial data, there is an obvious imbalance between normal records and abnormal. In this situation, the experimental results cannot be objectively evaluated only by the traditional metrics, such as precision, recall, and accuracy. This paper proposes an AutoEncoder enhanced LightGBM method for credit card detection. The method inherits the advantages of each component, using an AutoEncoder for feature reconstruction on the dataset, and integrating the LightGBM algorithm for improving the GBDT (Gradient Boosting Decison Tree) to detect abnormal data more accurately and efficiently. Besides the traditional evaluation metrics, F-measure, area under curve (AUC), Matthew’s correlation coefficient (MCC), and balanced classification rate (BCR) are also adopted as the evaluation metrics. Two financial datasets were used to validate the performance and robustness of the proposed model. Results obtained from the credit card fraud dataset containing 31 features indicate that our model significantly outperforms other models with a recall of 94.85%, representing a 10.70% improvement compared to the best detection performance model with a recall of only 86%. Additionally, our model’s BCR score is also significantly better than other models, with a BCR score of 97%, as opposed to the best detection performance model’s BCR score of 92%, representing a 5% improvement by our model. Various sampling methods and model combinations were considered in this study. It was found that the SMOTE algorithm combined with the proposed model produced the best results, with an AUC value of 96.83% and an F-measure score of 80.27%. The Santander bank transaction record dataset is a high dimensional large dataset containing 200 features. Experimental results on this dataset reveal that compared to other models, our model significantly improved recall and F-measure results, raising the recall to 94.14% and the F-measure score by 11.51%, surpassing the second-best-performing model. Overall, these findings demonstrate the robustness and superiority of our model in detecting fraudulent transactions and highlight the effectiveness of the SMOTE algorithm in combination with the proposed model.

Comparing Human Performance on Target Localization in Near Infrared and Long Wave Infrared for Cluttered Environments.

In the context of rapid advancements in AI, the accuracies and speeds among various AI models and methods are often compared. However, a basic question is rarely asked: is AI better than humans, and if so, under what conditions? This paper investigates human ability to detect distant landmark targets under cluttered surroundings such as buildings, trees, and clouds in NIR and LWIR images, aiming to facilitate AI object detection performance analysis. Our investigation employs perception tests and a human performance model to analyze object detection capabilities. The results reveal distinctive differences in NIR and LWIR detectability, showing that although LWIR performs less effectively at range, it offers superior robustness across various environmental conditions. Our findings suggest that AI could be particularly advantageous for object detection in LWIR as it outperform humans in terms of detection accuracy at a long range.

Clubs and events: Two models of performance in studies of sport and national solidarity

AbstractThis study distinguishes between two performances of solidarity—a performer–audience structure and a network structure—illustrated through a review of two case studies from the field of sport: a cross‐national survey of football fans and a historical study of French cycling clubs. Research on nationhood, sport and performance has centred mainly on public events and how they generate focused attention to foster mass solidarity. Scholars have overlooked an alternative model of performance situated in horizontal networks and interpersonal interactions staged in front of a third party, such as found in clubs. Distinguishing between these performance models in sport and beyond can further our understanding of the relations between identity and solidarity in studies of nationhood. Whereas in the former model, expressions of solidarity mainly reaffirm pre‐existing feelings of national identity; in the latter model, solidarity is performative: it does not simply represent social reality but serves to create it.

Linkages between service recovery system and customer justice perceptions: A multi-level model of employee service recovery performance.

Past research aimed at service recovery has focused on actions that are taken to retain customers, and the possibility of overcoming the mistakes of service delivery, though the multidimensional nature of the service recovery process has begun to move beyond the complaint handling process. In this paper, we identify the most important issues addressed in service recovery and present a framework for addressing them. We used an extensive integrative review methodology. Between 1988 and 2017 the number of articles in these journals was kept between 26 and 30. The majority took a business perspective to study services recovery, while a minority took an inter-disciplinarity perspective. The study's findings are expected to provide insights into the antecedents and consequences of service recovery performance, particularly how job resources and demands influence employee burnout and work engagement, which in turn affect service recovery performance and customer perceptions of justice. The research aims to contribute to the discourse on service recovery by building a multi-level model that explains the service recovery performance of frontline employees and the impact of the service recovery system on customer justice perceptions. Thus, this research is expected to contribute to robust conclusions on the antecedents and consequences of employee service recovery performance nested within the unit-level construct, i.e., the service recovery system.

A unified longitudinal trajectory dataset for automated vehicle

Automated Vehicles (AVs) promise significant advances in transportation. Critical to these improvements is understanding AVs’ longitudinal behavior, relying heavily on real-world trajectory data. Existing open-source trajectory datasets of AV, however, often fall short in refinement, reliability, and completeness, hindering effective performance metrics analysis and model development. This study addresses these challenges by creating a Unified longitudinal trajectory dataset for AVs (Ultra-AV) to analyze their microscopic longitudinal driving behaviors. This dataset compiles data from 14 distinct sources, encompassing various AV types, test sites, and experiment scenarios. We established a three-step data processing: 1. extraction of longitudinal trajectory data, 2. general data cleaning, and 3. data-specific cleaning to obtain the longitudinal trajectory data and car-following trajectory data. The validity of the processed data is affirmed through performance evaluations across safety, mobility, stability, and sustainability, along with an analysis of the relationships between variables in car-following models. Our work not only furnishes researchers with standardized data and metrics for longitudinal AV behavior studies but also sets guidelines for data collection and model development.