Machine learning with applications最新文献_第7页

Machine learning-based prediction of optimal antenatal care utilization among reproductive women in Nigeria 基于机器学习的尼日利亚育龄妇女最佳产前护理利用预测

Machine learning with applications Pub Date : 2025-06-27 DOI: 10.1016/j.mlwa.2025.100698

Jamilu Sani , Adeyemi Oluwagbemiga , Mohamed Mustaf Ahmed

{"title":"Machine learning-based prediction of optimal antenatal care utilization among reproductive women in Nigeria","authors":"Jamilu Sani , Adeyemi Oluwagbemiga , Mohamed Mustaf Ahmed","doi":"10.1016/j.mlwa.2025.100698","DOIUrl":"10.1016/j.mlwa.2025.100698","url":null,"abstract":"<div><h3>Background</h3><div>Despite global efforts, disparities in antenatal care (ANC) utilization persist in Nigeria, where maternal mortality remains alarmingly high (1047 deaths per 100,000 live births). Traditional statistical models often fall short in identifying complex non-linear relationships in population health data. Machine learning (ML) offers a promising alternative that uncovers hidden patterns and improves prediction accuracy.</div></div><div><h3>Methods</h3><div>This study used data from the 2018 Nigeria Demographic and Health Survey (NDHS), a nationally representative data set. After data preprocessing and feature selection, six supervised ML algorithms—Logistic Regression, Support Vector Machine, K-Nearest Neighbors, Decision Tree, Random Forest, and XGBoost—were applied using Python 3.9. The model performance was evaluated using accuracy, precision, recall, F1-score, and area under the receiver operating characteristic curve (AUROC). Feature importance was assessed using permutation importance and Gini impurity score.</div></div><div><h3>Results</h3><div>Among all models, Random Forest achieved the best performance, with 90 % accuracy, 0.90 precision and recall, an F1-score of 0.91, and an AUROC of 0.90. Permutation and Gini importance analyses identified the place of delivery, region, residence, and educational level as the most influential predictors. Other moderately important features included distance to health facilities, husband’s occupation, number of births, and healthcare decision-making autonomy—factors not highlighted by traditional statistical approaches.</div></div><div><h3>Conclusion</h3><div>Machine learning, particularly Random Forest, demonstrated strong predictive power in identifying the key determinants of ANC utilization. These findings highlight the potential of ML to inform targeted maternal health interventions and improve outcomes in low-resource settings, such as Nigeria.</div></div>","PeriodicalId":74093,"journal":{"name":"Machine learning with applications","volume":"21 ","pages":"Article 100698"},"PeriodicalIF":0.0,"publicationDate":"2025-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144523203","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Enhanced credit risk prediction using deep learning and SMOTE-ENN resampling 利用深度学习和SMOTE-ENN重采样增强信用风险预测

Machine learning with applications Pub Date : 2025-06-27 DOI: 10.1016/j.mlwa.2025.100692

Idowu Aruleba, Yanxia Sun

{"title":"Enhanced credit risk prediction using deep learning and SMOTE-ENN resampling","authors":"Idowu Aruleba, Yanxia Sun","doi":"10.1016/j.mlwa.2025.100692","DOIUrl":"10.1016/j.mlwa.2025.100692","url":null,"abstract":"<div><div>Credit risk prediction is a vital task in financial services, ensuring that institutions can manage their lending risks effectively. This study investigates the effectiveness of deep learning (DL) models for credit risk prediction, with a focus on addressing the challenge of class imbalance and the black box nature of these models using the Synthetic Minority Over-sampling Technique - Edited Nearest Neighbor (SMOTE-ENN) resampling method and Shapley Additive Explanations (SHAP), respectively. The study compares the performance of various DL architectures, including Convolutional Neural Networks (CNN), Long Short-Term Memory networks (LSTM), Gated Recurrent Units (GRU), and Graph Neural Networks (GNN), on two real-world datasets: the Australian and German credit datasets. The findings reveal that the GRU model, enhanced with SMOTE-ENN resampling, outperforms other models in terms of accuracy, sensitivity, and specificity. The superior performance of the GRU-SMOTE-ENN model demonstrates its potential as a robust deep learning technique for financial institutions to enhance credit risk assessment. Additionally, the study demonstrates how the integration of SHAP values significantly improves the interpretability of deep learning models, making them more transparent and trustworthy for stakeholders.</div></div>","PeriodicalId":74093,"journal":{"name":"Machine learning with applications","volume":"21 ","pages":"Article 100692"},"PeriodicalIF":0.0,"publicationDate":"2025-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144510825","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Performance benchmarking of multimodal data-driven approaches in industrial settings 工业环境中多模式数据驱动方法的性能基准测试

Machine learning with applications Pub Date : 2025-06-25 DOI: 10.1016/j.mlwa.2025.100691

Diyar Altinses, Andreas Schwung

引用次数: 0

Hierarchical data modeling: A systematic comparison of statistical, tree-based, and neural network approaches 分层数据建模：统计、基于树和神经网络方法的系统比较

Machine learning with applications Pub Date : 2025-06-24 DOI: 10.1016/j.mlwa.2025.100688

Marzieh Amiri Shahbazi, Nasibeh Azadeh-Fard

{"title":"Hierarchical data modeling: A systematic comparison of statistical, tree-based, and neural network approaches","authors":"Marzieh Amiri Shahbazi, Nasibeh Azadeh-Fard","doi":"10.1016/j.mlwa.2025.100688","DOIUrl":"10.1016/j.mlwa.2025.100688","url":null,"abstract":"<div><div>Hierarchical modeling approaches have evolved significantly, yet comprehensive comparisons between fundamentally different methodological paradigms remain limited. This research presents a systematic comparative analysis of three distinct hierarchical modeling approaches: statistical (Hierarchical Mixed Model), tree-based (Hierarchical Random Forest), and neural (Hierarchical Neural Network). Based on the 2019 National Inpatient Sample — comprising more than seven million records from 4568 hospitals across four U.S. regions — the models were assessed for their ability to predict length of stay at the patient, hospital, and regional levels. The evaluation framework integrated quantitative metrics and qualitative factors, including analyses across varying sample sizes, simplified hierarchies, and a separate intensive-care dataset. Results demonstrate that tree-based approaches consistently outperform alternatives in predictive accuracy and explanation of variance while maintaining computational efficiency. These performance patterns remain generally consistent across sample sizes, simplified hierarchies, and the external dataset. Neural approaches excel at capturing group-level distinctions but require substantial computational resources and exhibit prediction bias. Statistical approaches offer rapid inference and interpretability but underperform in accuracy at intermediate hierarchical levels. Each model exhibits distinctive hierarchical information processing: neural models favor bottom-up flow, statistical models emphasize top-down constraints, and tree-based models achieve balanced integration. This research establishes practical guidelines for selecting appropriate hierarchical modeling approaches based on data characteristics, computational constraints, and analytical requirements, thereby advancing understanding of fundamental trade-offs in multilevel analysis.</div></div>","PeriodicalId":74093,"journal":{"name":"Machine learning with applications","volume":"21 ","pages":"Article 100688"},"PeriodicalIF":0.0,"publicationDate":"2025-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144480060","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Embedded feature selection using dual-network architecture 采用双网络架构的嵌入式特征选择

Machine learning with applications Pub Date : 2025-06-18 DOI: 10.1016/j.mlwa.2025.100672

Abderrahim Abbassi, Arved Dörpinghaus, Niklas Römgens, Tanja Grießmann, Raimund Rolfes

{"title":"Embedded feature selection using dual-network architecture","authors":"Abderrahim Abbassi, Arved Dörpinghaus, Niklas Römgens, Tanja Grießmann, Raimund Rolfes","doi":"10.1016/j.mlwa.2025.100672","DOIUrl":"10.1016/j.mlwa.2025.100672","url":null,"abstract":"<div><div>Feature selection is essential for eliminating noise, reducing redundancy, simplifying computational complexity, and lowering data collection and processing costs. However, existing methods often face challenges due to the complexity of feature interdependencies, uncertainty regarding the exact number of relevant features, and the need for hyperparameter optimization, which increases methodological complexity.</div><div>This research proposes a novel dual-network architecture for feature selection that addresses these issues. The architecture consists of a task model and a selection model. First, redundant features are fed into the selection model, which generates a binary mask aligned with the input feature dimensions. This mask is applied to a shifted version of the original features, serving as input to the task model. The task model then uses the selected features to perform the target supervised task. Simultaneously, the selection model aims to minimize the cumulative value of the mask, thus selecting the most relevant features with minimal impact on the task model’s performance.</div><div>The method is evaluated using benchmark and synthetic datasets across different supervised tasks. Comparative evaluation with state-of-the-art techniques demonstrates that the proposed approach exhibits superior or competitive feature selection capabilities, achieving a reduction of 90% or more in feature count. This is particularly notable in the presence of non-linear feature interdependencies. The key advantages of the proposed method are its ability to self-determine the number of relevant features needed for the supervised task and its simplicity, requiring the pre-definition of only a single hyperparameter, for which an estimation approach is suggested.</div></div>","PeriodicalId":74093,"journal":{"name":"Machine learning with applications","volume":"21 ","pages":"Article 100672"},"PeriodicalIF":0.0,"publicationDate":"2025-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144489970","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

VICCA: Visual interpretation and comprehension of chest X-ray anomalies in generated report without human feedback VICCA：无人工反馈的生成报告中胸部x线异常的视觉解释和理解

Machine learning with applications Pub Date : 2025-06-18 DOI: 10.1016/j.mlwa.2025.100684

Sayeh Gholipour Picha, Dawood Al Chanti, Alice Caplier

{"title":"VICCA: Visual interpretation and comprehension of chest X-ray anomalies in generated report without human feedback","authors":"Sayeh Gholipour Picha, Dawood Al Chanti, Alice Caplier","doi":"10.1016/j.mlwa.2025.100684","DOIUrl":"10.1016/j.mlwa.2025.100684","url":null,"abstract":"<div><div>As artificial intelligence (AI) becomes increasingly central to healthcare, the demand for explainable and trustworthy models is paramount. Current report generation systems for chest X-rays (CXR) often lack mechanisms for validating outputs without expert oversight, raising concerns about reliability and interpretability. To address these challenges, we propose a novel multimodal framework designed to enhance the semantic alignment between text and image context and the localization accuracy of pathologies within images and reports for AI-generated medical reports. Our framework integrates two key modules: a Phrase Grounding Model, which identifies and localizes pathologies in CXR images based on textual prompts, and a Text-to-Image Diffusion Module, which generates synthetic CXR images from prompts while preserving anatomical fidelity. By comparing features between the original and generated images, we introduce a dual-scoring system: one score quantifies localization accuracy, while the other evaluates semantic consistency between text and image features. Our approach significantly outperforms existing methods in pathology localization, achieving an 8% improvement in Intersection over Union score. It also surpasses state-of-the-art methods in CXR text-to-image generation, with a 1% gain in similarity metrics. Additionally, the integration of phrase grounding with diffusion models, coupled with the dual-scoring evaluation system, provides a robust mechanism for validating report quality, paving the way for more reliable and transparent AI in medical imaging.</div></div>","PeriodicalId":74093,"journal":{"name":"Machine learning with applications","volume":"21 ","pages":"Article 100684"},"PeriodicalIF":0.0,"publicationDate":"2025-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144366295","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

AdVision: An efficient and effective deep learning based advertisement detector for printed media AdVision：一个高效的基于深度学习的印刷媒体广告检测器

Machine learning with applications Pub Date : 2025-06-18 DOI: 10.1016/j.mlwa.2025.100686

Faeze Zakaryapour Sayyad , Irida Shallari , Seyed Jalaleddin Mousavirad , Mattias O’Nils , Faisal Z. Qureshi

{"title":"AdVision: An efficient and effective deep learning based advertisement detector for printed media","authors":"Faeze Zakaryapour Sayyad , Irida Shallari , Seyed Jalaleddin Mousavirad , Mattias O’Nils , Faisal Z. Qureshi","doi":"10.1016/j.mlwa.2025.100686","DOIUrl":"10.1016/j.mlwa.2025.100686","url":null,"abstract":"<div><div>Automated advertisement detection in newspapers is a challenging task due to the diversity in print layouts, formats, and design styles. This task has critical applications in media monitoring, content analysis, and advertising analytics. To address these challenges, we introduce AdVision, a deep-learning-based solution that treats advertisements as unique visual objects. We provide a comparative study of various detection architectures, including one-stage, two-stage, and transformer-based detectors, to identify the most effective approach for detecting advertisements. Our results are validated through extensive experiments conducted under different conditions and metrics. Newspapers from four different countries — Denmark, Norway, Sweden, and the UK — were selected to demonstrate the variety of languages and print formats. Additionally, we conduct a cross-analysis to show how training on one language can generalize to another. To enhance the explainability of our results, we employ GradCAM++ (Chattopadhay et al., 2018) heatmaps. Our experiments demonstrate that the YOLOv8 model achieves superior performance, balancing high precision and recall with minimal inference latency, making it particularly suitable for high-throughput advertisement detection.</div></div>","PeriodicalId":74093,"journal":{"name":"Machine learning with applications","volume":"21 ","pages":"Article 100686"},"PeriodicalIF":0.0,"publicationDate":"2025-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144338897","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Artificial neural networks and support vector machines for more accurate cost estimation in underground mining: A contractor's viewpoint 基于人工神经网络和支持向量机的地下采矿成本估算：一个承包商的观点

Machine learning with applications Pub Date : 2025-06-15 DOI: 10.1016/j.mlwa.2025.100689

Juan Camilo García Vásquez, Mustafa Kumral

{"title":"Artificial neural networks and support vector machines for more accurate cost estimation in underground mining: A contractor's viewpoint","authors":"Juan Camilo García Vásquez, Mustafa Kumral","doi":"10.1016/j.mlwa.2025.100689","DOIUrl":"10.1016/j.mlwa.2025.100689","url":null,"abstract":"<div><div>Accurate cost estimation is crucial in effective decision-making and evaluation in underground mining projects. Machine learning techniques have shown enormous potential in enhancing cost estimation accuracy in various industries. This study harnesses artificial neural networks (ANN) and Support Vector Machines (SVM) to estimate operating costs in underground mining. Special emphasis is placed on cost estimation from a contractor’s perspective. Mining contractors are sensitive to deviations from the estimated costs because slight deviations may result in losing a contract bid or financial loss in an awarded project. The proposed approach can help contractors make more informed decisions and improve project management. Comprehensive data containing various parameters that impact the cost of underground mining projects, such as equipment type utilization, rock type, and cross-sectional area, were collected. This dataset was used to train and evaluate ANN and SVM models that provide more accurate cost estimation for underground mining projects. The best model achieved a mean average percentage error (MAPE) of 5.31 % for the ANN model and 3.05 % for the SVM model, outperforming traditional cost estimation methods. This study demonstrates the potential of machine learning in enhancing the performance of the cost estimation process.</div></div>","PeriodicalId":74093,"journal":{"name":"Machine learning with applications","volume":"21 ","pages":"Article 100689"},"PeriodicalIF":0.0,"publicationDate":"2025-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144338896","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Multimodal bearing fault classification under variable conditions: A 1D CNN with transfer learning 可变条件下的多模态轴承故障分类：带有迁移学习的1D CNN

Machine learning with applications Pub Date : 2025-06-14 DOI: 10.1016/j.mlwa.2025.100682

Tasfiq E. Alam, Md Manjurul Ahsan, Shivakumar Raman

{"title":"Multimodal bearing fault classification under variable conditions: A 1D CNN with transfer learning","authors":"Tasfiq E. Alam, Md Manjurul Ahsan, Shivakumar Raman","doi":"10.1016/j.mlwa.2025.100682","DOIUrl":"10.1016/j.mlwa.2025.100682","url":null,"abstract":"<div><div>Bearings play an integral role in ensuring the reliability and efficiency of rotating machinery — reducing friction and handling critical loads. Bearing failures that constitute up to 90% of mechanical faults highlight the imperative need for reliable condition monitoring and fault detection. This study proposes a multimodal bearing fault classification approach that relies on vibration and motor phase current signals within a one-dimensional convolutional neural network (1D CNN) framework. The method fuses features from multiple signals to enhance the accuracy of fault detection. Under the baseline condition (1500 rpm, 0.7 Nm load torque, and 1000 N radial force), the model reaches an accuracy of 96% with addition of L2 regularization. In addition, the model demonstrates robust performance across three distinct operating conditions by employing transfer learning (TL) strategies. Among the tested TL variants, the approach that preserves parameters up to the first max-pool layer and then adjusts subsequent layers achieves the highest performance. While this approach attains excellent accuracy across varied conditions, it requires more computational time due to its greater number of trainable parameters. To address resource constraints, less computationally intensive models offer feasible trade-offs, albeit at a slight accuracy cost. Overall, this multimodal 1D CNN framework with late fusion and TL strategies lays a foundation for more accurate, adaptable, and efficient bearing fault classification in industrial environments with variable operating conditions.</div></div>","PeriodicalId":74093,"journal":{"name":"Machine learning with applications","volume":"21 ","pages":"Article 100682"},"PeriodicalIF":0.0,"publicationDate":"2025-06-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144297023","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

AI-driven adaptive mesh refinement for thermal–hydraulic simulations in nuclear reactors 核反应堆热工模拟中人工智能驱动的自适应网格细化

Machine learning with applications Pub Date : 2025-06-14 DOI: 10.1016/j.mlwa.2025.100670

Shuai Ren, Xue Miao, Huizhao Li, Lingyu Dong, Dandan Chen

{"title":"AI-driven adaptive mesh refinement for thermal–hydraulic simulations in nuclear reactors","authors":"Shuai Ren, Xue Miao, Huizhao Li, Lingyu Dong, Dandan Chen","doi":"10.1016/j.mlwa.2025.100670","DOIUrl":"10.1016/j.mlwa.2025.100670","url":null,"abstract":"<div><div>The meshing of complex flow channels is the most time-consuming part of large-scale thermal–hydraulic simulations in nuclear reactors and often struggles to converge. Machine learning is employed to guide the optimization of the wire-wrapped fuel rod channel meshing, which has been successfully applied to large-scale fluid simulations. The main contributions of this paper are as follows: (1) A novel adaptive meshing technology based on ”adaptive meshing + machine learning algorithms” is proposed and successfully applied to predict sensitive channel meshes and achieve automatic refinement in nuclear reactors; (2) By comparing the channel mesh models before and after optimization, mesh quality was improved while maintaining the boundary integrity of the initial channel mesh model; (3) Based on the mesh refinement algorithm, a mesh refinement tool was developed and successfully coupled with classical thermal–hydraulic simulation software, enabling the thermal–hydraulic computation of a two-dimensional axial wire-wrapped flow channel in a nuclear reactor; (4) The performance of the coupled model was evaluated, demonstrating a relative speedup of 144.54 and parallel efficiency of 56.4% when scaled to 256 cores. Since this algorithm is developed based on the general characteristics of physical object discretization simulations, it holds the potential for cross-domain applications.</div></div>","PeriodicalId":74093,"journal":{"name":"Machine learning with applications","volume":"21 ","pages":"Article 100670"},"PeriodicalIF":0.0,"publicationDate":"2025-06-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144313831","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0