Array最新文献 - Book学术

Discriminative local affine-hull clustering for high-dimensional data 高维数据的判别局部仿射-船体聚类

IF 2.3

Array Pub Date : 2025-07-23 DOI: 10.1016/j.array.2025.100465

Yu-Feng Yu , Jiali Luo , Xuanyi Chen , Yingchao Cheng , Yulin He , Joshua Zhexue Huang

{"title":"Discriminative local affine-hull clustering for high-dimensional data","authors":"Yu-Feng Yu , Jiali Luo , Xuanyi Chen , Yingchao Cheng , Yulin He , Joshua Zhexue Huang","doi":"10.1016/j.array.2025.100465","DOIUrl":"10.1016/j.array.2025.100465","url":null,"abstract":"<div><div>Clustering high-dimensional data presents a critical technical challenge due to the curse of dimensionality, feature redundancy, and sensitivity to noise—issues that significantly degrade clustering accuracy in applications such as gene expression analysis, image recognition, and anomaly detection. Existing solutions often rely on dimensionality reduction techniques that risk discarding discriminative features, or on deep learning methods that require large-scale training data and suffer from poor interpretability. To address these limitations, this study proposes a novel discriminative subspace clustering algorithm that avoids traditional dimensionality reduction and instead operates directly in the high-dimensional space. Our method partitions the sample space into multiple local affine hulls and introduces a discriminative geometric distance metric that accounts for both relevant and irrelevant subspaces. Specifically, the model measures the ratio between a query sample’s proximity to its class-specific affine hull and its distance from unrelated class subspaces. This dual-space modeling improves both intra-class compactness and inter-class separation. To ensure computational efficiency, we reformulate distance calculations as matrix multiplications and leverage SVD for subspace projection, enabling scalable performance across large datasets. Extensive experiments on seven benchmark datasets demonstrate that the proposed method consistently outperforms state-of-the-art clustering algorithms. It achieves up to 92.60% accuracy on MNIST and maintains high robustness on sparse and noisy data, validating its effectiveness for high-dimensional clustering tasks. This work contributes a geometrically interpretable and computationally efficient framework that closes a long-standing gap in unsupervised learning under high-dimensional constraints.</div></div>","PeriodicalId":8417,"journal":{"name":"Array","volume":"27 ","pages":"Article 100465"},"PeriodicalIF":2.3,"publicationDate":"2025-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144714345","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

NEDL-GCP: A nested ensemble deep learning model for Gynecological cancer risk prediction NEDL-GCP：用于妇科癌症风险预测的嵌套集成深度学习模型

IF 2.3

Array Pub Date : 2025-07-23 DOI: 10.1016/j.array.2025.100468

Kamal Berahmand , Xujuan Zhou , Yuefeng Li , Raj Gururajan , Prabal Datta Barua , U Rajendra Acharya , Srinivas Kondalsamy Chennakesavan

{"title":"NEDL-GCP: A nested ensemble deep learning model for Gynecological cancer risk prediction","authors":"Kamal Berahmand , Xujuan Zhou , Yuefeng Li , Raj Gururajan , Prabal Datta Barua , U Rajendra Acharya , Srinivas Kondalsamy Chennakesavan","doi":"10.1016/j.array.2025.100468","DOIUrl":"10.1016/j.array.2025.100468","url":null,"abstract":"<div><div>Gynecological cancer remains a critical global health concern, where early detection significantly improves patient outcomes. Despite advances in deep learning for medical diagnostics, existing models often struggle with feature redundancy, lack of generalizability, and suboptimal integration of diverse feature representations, limiting their effectiveness in clinical applications. In this study, we present NEDL-GCP, a Nested Ensemble Deep Learning model for Gynecological Cancer Risk Prediction, which uses a hierarchical ensemble framework to improve the accuracy of the classification. NEDL-GCP integrates CNNs, RNNs, and SVMs as base learners, extracting diverse feature representations, while a meta-classifier combining J48 and Stochastic Gradient Descent (SGD) refines predictions. Evaluated on the Herlev and SIPaKMeD Pap Smear datasets, NEDL-GCP achieved state-of-the-art accuracy scores of 99.1% and 98.5%, outperforming existing methods. These results demonstrate the robustness and reliability of the model, making it a valuable tool for the early detection of cervical cancer. By enhancing diagnostic accuracy and optimizing clinical workflows, NEDL-GCP supports timely decision-making, ultimately improving patient care.</div></div>","PeriodicalId":8417,"journal":{"name":"Array","volume":"27 ","pages":"Article 100468"},"PeriodicalIF":2.3,"publicationDate":"2025-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144713736","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

Prediction of psychophysiological perception of driver in road tunnel using particle swarm optimization improved stacked denoising autoencoder model 基于粒子群优化改进的叠加去噪自编码器模型的公路隧道驾驶员心理生理感知预测

IF 2.3

Array Pub Date : 2025-07-22 DOI: 10.1016/j.array.2025.100475

Can Qin , Bo Liang , Jia'an Niu , Jinghang Xiao , Shuangkai Zhu , Haonan Long

{"title":"Prediction of psychophysiological perception of driver in road tunnel using particle swarm optimization improved stacked denoising autoencoder model","authors":"Can Qin , Bo Liang , Jia'an Niu , Jinghang Xiao , Shuangkai Zhu , Haonan Long","doi":"10.1016/j.array.2025.100475","DOIUrl":"10.1016/j.array.2025.100475","url":null,"abstract":"<div><div>Technical limitations of test equipment, changes in test environment, and jerks of test vehicles under the lighting environment of highway tunnels can lead to the appearance of abnormal psychophysiological data, which affects the data quality and the subsequent prediction and analysis. In this study, the physical quantities of lighting environment and the psychophysiological quantities (heart rate, pupil area, recognition distance and reaction time) of drivers were collected, and the information representation of physical quantities affecting the perception ability of psychophysiological quantities were evaluated and screened in terms of importance of variables by correlation analysis method and LASSO-CV regression method. Based on the screened key physical quantities, particle swarm optimization (PSO) was employed to set the hyper-parameters for stacked denoising autoencoder (SDAE), and the prediction results of the PSO-SDAE model were compared with other network methods, and then the partial dependence plot was used to further explore the intrinsic mechanism of information representation for physical quantities. The results show that the proposed PSO-SDAE model can effectively achieves the reasonable configuration of the SDAE network parameters, and clean abnormal data by mining the hidden information and structural features of normal data. The PSO-SDAE model has an excellent prediction accuracy, stability and cleaning effect when facing different scales and types of normal or abnormal data for psychophysiological quantities.</div></div>","PeriodicalId":8417,"journal":{"name":"Array","volume":"27 ","pages":"Article 100475"},"PeriodicalIF":2.3,"publicationDate":"2025-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144702723","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

A reinforcement learning based fuzzing technique for binary programs vulnerabilities detection 基于强化学习的模糊检测技术在二进制程序漏洞检测中的应用

IF 2.3

Array Pub Date : 2025-07-21 DOI: 10.1016/j.array.2025.100458

Guoyan Cao , Yanhui Ma , Mengjiao Geng

引用次数: 0

Humidity and temperature monitoring using Raspberry Pi via RS232 networking 湿度和温度监测使用树莓派通过RS232网络

IF 2.3

Array Pub Date : 2025-07-19 DOI: 10.1016/j.array.2025.100464

Pradhyut Rajkumar

引用次数: 0

A framework for post-prognosis decision-making utilizing deep reinforcement learning considering imperfect maintenance decisions and Value of Information 考虑不完善维护决策和信息价值的深度强化学习后预后决策框架

IF 2.3

Array Pub Date : 2025-07-19 DOI: 10.1016/j.array.2025.100454

P. Komninos, D. Zarouchas

{"title":"A framework for post-prognosis decision-making utilizing deep reinforcement learning considering imperfect maintenance decisions and Value of Information","authors":"P. Komninos, D. Zarouchas","doi":"10.1016/j.array.2025.100454","DOIUrl":"10.1016/j.array.2025.100454","url":null,"abstract":"<div><div>The digitalization era has introduced an abundance of data that can be harnessed to monitor and predict the health of structures. This paper presents a comprehensive framework for post-prognosis decision-making that utilizes deep reinforcement learning (DRL) to manage maintenance decisions on multi-component systems subject to imperfect repairs. The proposed framework integrates raw sensory data acquisition, feature extraction, prognostics, imperfect repair modeling, and decision-making. This integration considers all these tasks independent, promoting flexibility and paving the way for more advanced and adaptable maintenance solutions in real-world applications. The framework’s effectiveness is demonstrated through a case study involving tension-tension fatigue experiments on open-hole aluminum coupons representing multiple dependent components, where the ability to make stochastic RUL estimations and schedule maintenance actions is evaluated. The results demonstrate that the framework can effectively extend the lifecycle of the system while accommodating uncertainties in maintenance actions. This work utilizes the Value of Information to choose the optimal times to acquire new data, resulting in computational efficiency and significant resource savings. Finally, it emphasizes the importance of decomposing uncertainty into epistemic and aleatoric to convert the total uncertainty into decision probabilities over the chosen actions, ensuring reliability and enhancing the interpretability of the DRL model.</div></div>","PeriodicalId":8417,"journal":{"name":"Array","volume":"27 ","pages":"Article 100454"},"PeriodicalIF":2.3,"publicationDate":"2025-07-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144678953","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

Taxonomy, challenges, and future directions for AI-driven industrial cooling systems 人工智能驱动的工业冷却系统的分类、挑战和未来方向

IF 2.3

Array Pub Date : 2025-07-17 DOI: 10.1016/j.array.2025.100448

Md Mohsin Kabir, Shahina Begum, Shaibal Barua, Mobyen Uddin Ahmed

{"title":"Taxonomy, challenges, and future directions for AI-driven industrial cooling systems","authors":"Md Mohsin Kabir, Shahina Begum, Shaibal Barua, Mobyen Uddin Ahmed","doi":"10.1016/j.array.2025.100448","DOIUrl":"10.1016/j.array.2025.100448","url":null,"abstract":"<div><div>The efficiency and reliability of industrial cooling systems are critical for sectors such as energy systems, electronics manufacturing, and data centers. Traditional cooling systems rely on reactive maintenance, leading to increased downtime, energy consumption, and operating costs. Recent advances in artificial intelligence (AI), including machine learning (ML), deep learning (DL), and physics-informed neural networks (PINNs), have enabled proactive fault diagnosis and predictive maintenance in industrial cooling systems, significantly reducing energy use and improving operational reliability. However, current AI applications face challenges, such as limited access to quality datasets, computational complexity, integration with legacy systems, and model scalability. This paper systematically addresses these gaps by providing a detailed taxonomy of AI-driven cooling system diagnostics, categorizing state-of-the-art methods, and identifying critical research challenges. Our main contribution is a structured taxonomy that integrates ML, DL, and PINNs, offering a clear framework for analyzing current practices and potential improvements. The paper highlights critical insights across 138 reviewed studies, emphasizing the transformative role of hybrid AI frameworks in diagnostics, including use cases in HVAC, data centers, and thermal imaging. Notably, the integration of ML, DL, and PINNs has been shown to improve fault detection accuracy, energy efficiency, and model interpretability, paving the way for scalable, real-time deployments.</div></div>","PeriodicalId":8417,"journal":{"name":"Array","volume":"27 ","pages":"Article 100448"},"PeriodicalIF":2.3,"publicationDate":"2025-07-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144686413","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

Cobb angle prediction for adolescent idiopathic scoliosis via an explainable machine learning model 通过可解释的机器学习模型预测青少年特发性脊柱侧凸的Cobb角

IF 2.3

Array Pub Date : 2025-07-17 DOI: 10.1016/j.array.2025.100455

Yu Ding , Bin Li , Xiaoyong Guo

{"title":"Cobb angle prediction for adolescent idiopathic scoliosis via an explainable machine learning model","authors":"Yu Ding , Bin Li , Xiaoyong Guo","doi":"10.1016/j.array.2025.100455","DOIUrl":"10.1016/j.array.2025.100455","url":null,"abstract":"<div><div>This study aims to build an accurate and interpretable machine learning model capable of adolescent idiopathic scoliosis prognostication. A tree-based gradient boosting machine is incorporated with a recently proposed Shapley-value-based explanation method-TreeExplainer. Anthropometric training data are collected from a public orthopedics clinic, and each instance is characterized by nine features with a prediction target. We adopt a transfer-learning strategy that takes advantage of the additive property of tree-based gradient boosting, allowing a gradient boosting machine regressor to be trained with limited labeled examples. Cross-validation estimation shows a satisfactory performance for predicting future spine curvature (Cobb angle). The root mean square error (<span><math><msup><mrow></mrow><mrow><mo>∘</mo></mrow></msup></math></span>), the mean absolute percentage error (<span><math><msup><mrow></mrow><mrow><mo>∘</mo></mrow></msup></math></span>), and the Pearson correlation coefficient are 3.69 ± 1.23, 2.81 ± 1.69, and 0.92 ± 0.01, respectively. Moreover, the overfitting has been largely removed, and the model may be generalized well to new patients. A well-trained model is taken as the input to the TreeExplainer. The output of the TreeExplainer provides us a richer understanding that demonstrates how a feature’s value impacts the model’s prediction for every instance. The patterns identified can substantially improve the human-artificial intelligence collaboration in the clinical management of patients with adolescent idiopathic scoliosis by preventing serious scoliosis progression and reducing healthcare costs.</div></div>","PeriodicalId":8417,"journal":{"name":"Array","volume":"27 ","pages":"Article 100455"},"PeriodicalIF":2.3,"publicationDate":"2025-07-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144679060","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

Forecasting loan, deferred rate and customer segmentation in banking industry: A computational intelligence approach 银行业贷款预测、递延利率和客户细分：一种计算智能方法

IF 2.3

Array Pub Date : 2025-07-17 DOI: 10.1016/j.array.2025.100460

Mahtab Vasheghani, Ebrahim Nazari Farokhi, Behrooz Dolatshah

{"title":"Forecasting loan, deferred rate and customer segmentation in banking industry: A computational intelligence approach","authors":"Mahtab Vasheghani, Ebrahim Nazari Farokhi, Behrooz Dolatshah","doi":"10.1016/j.array.2025.100460","DOIUrl":"10.1016/j.array.2025.100460","url":null,"abstract":"<div><div>Accurate loan default prediction and customer segmentation are critical challenges in the banking industry. This study proposes a novel hybrid model integrating Multi-Layer Perceptron (MLP) neural networks with Self-Adaptive Particle Swarm Optimization (PSO) and Genetic Algorithm (GA) frameworks. GA handles feature selection, while PSO optimizes MLP hyperparameters (e.g., learning rate, neurons, activation functions). The model dynamically enhances classification accuracy and resilience, particularly for imbalanced datasets in loan default prediction. Using real-world data from Sina Bank, the system outperforms Logistic Regression, Decision Trees, and Random Forests. The GA-PSO optimization process, which integrates both PSO and GA to optimize the MLP model's parameters, plays a crucial role in enhancing the accuracy and scalability of the system. Specifically, the GA-PSO-MLP model achieves a 15 % higher classification accuracy than Logistic Regression, a 12 % improvement over Decision Trees, and an 8 % gain over Random Forests. Additionally, false positive rates are reduced by 20 %, and mean squared error (MSE) is lowered by 18 %. The F1-score of the proposed model is 92.3 %, compared to 79.8 % (Logistic Regression), 81.5 % (Decision Trees), and 85.2 % (Random Forests), further highlighting its advantage in handling imbalanced datasets. Extensive numerical validation and sensitivity analysis further highlight the model's effectiveness in delivering actionable insights that enhance customer management strategies and mitigate financial risks. This research makes a substantial contribution to the application of machine learning in banking, facilitating more accurate data-driven decision-making and more robust risk management practices.</div></div>","PeriodicalId":8417,"journal":{"name":"Array","volume":"27 ","pages":"Article 100460"},"PeriodicalIF":2.3,"publicationDate":"2025-07-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144678951","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

The future of libraries: Integrating pepper and computer vision for smart assistance 图书馆的未来：整合胡椒和计算机视觉的智能辅助

IF 2.3

Array Pub Date : 2025-07-17 DOI: 10.1016/j.array.2025.100469

Claire Trinquet, Deepti Mishra , Akshara Pande

{"title":"The future of libraries: Integrating pepper and computer vision for smart assistance","authors":"Claire Trinquet, Deepti Mishra , Akshara Pande","doi":"10.1016/j.array.2025.100469","DOIUrl":"10.1016/j.array.2025.100469","url":null,"abstract":"<div><div>In recent decades, the utilization of social robots in our daily lives has increased, but they are different from robots designed for libraries. On the one hand, library robots cannot establish social interactions, while social robots lack the necessary sensors to identify books. The present study aims to integrate the social robot Pepper's camera with computer vision techniques to enable Pepper to read the titles of books in front of it. This involves two main steps. The first step is to detect objects, i.e. books, from the scene. Thereafter, the titles of the books need to be read from the previous step. To achieve the first objective, two object detection models, YOLOv4 and YOLOv9, were employed. To accomplish the second goal, three OCR models —EasyOCR, Pytesseract, and Keras-OCR — were used. The results indicate that with the YOLOv9 model, all books were detected, whereas with the YOLOv4 model, 94 % books were identified. The findings of the present study suggest that when the YOLOv4 model and YOLOv9 were applied, EasyOCR performed well at a distance of 50 cm with a resolution of 3. Although the results of the OCR do not match perfectly with the written text on the books, the error rate is quite low for recognition by humans and computers. Therefore, there is a need to employ more advanced object detection and OCR technologies in future work.</div></div>","PeriodicalId":8417,"journal":{"name":"Array","volume":"27 ","pages":"Article 100469"},"PeriodicalIF":2.3,"publicationDate":"2025-07-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144663694","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