Intelligent Systems with Applications最新文献

{"title":"An adaptive chaotic league championship algorithm for solving global optimization and engineering design problems","authors":"Tanachapong Wangkhamhan,&nbsp;Jatsada Singthongchai","doi":"10.1016/j.iswa.2025.200511","DOIUrl":"10.1016/j.iswa.2025.200511","url":null,"abstract":"<div><div>This paper introduces a novel approach to global numerical optimization through the development of an Adaptive Chaotic League Championship Algorithm (AC-LCA). Our methodology enhances the conventional League Championship Algorithm (LCA) by integrating an adaptive chaotic local search mechanism. This integration aims to improve the exploration and exploitation capabilities of the LCA, enabling it to effectively navigate complex search spaces and avoid premature convergence. Abundant experiments have been extensively executed on the well-known CEC2017 benchmark problem sets to validate the performance of AC-LCA. The results demonstrate significant improvements in convergence speed and solution accuracy over traditional LCA and several other state-of-the-art optimization algorithms. Notably, the adaptive chaotic component plays a critical role in fine-tuning the search process, contributing to the robustness and efficiency of the algorithm. The paper also investigates the application of AC-LCA to a set of five famous real-life engineering problems, showcasing its practicality and adaptability in diverse optimization scenarios. These applications further underline the algorithm's potential to address a wide range of complex optimization tasks, making it a valuable tool for researchers and practitioners in the field. Overall, the AC-LCA emerges as a promising new approach in global numerical optimization, offering a balance of innovative methodology and practical applicability.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"26 ","pages":"Article 200511"},"PeriodicalIF":0.0,"publicationDate":"2025-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143748710","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"FocusAugMix: A data augmentation method for enhancing Acute Lymphoblastic Leukemia classification","authors":"Tanzilal Mustaqim ,&nbsp;Chastine Fatichah ,&nbsp;Nanik Suciati ,&nbsp;Takashi Obi ,&nbsp;Joong-Sun Lee","doi":"10.1016/j.iswa.2025.200512","DOIUrl":"10.1016/j.iswa.2025.200512","url":null,"abstract":"<div><div>The detection of various subtypes of Acute Lymphoblastic Leukemia (ALL) is crucial for precise medical identification, even though it is often hindered by the diverse appearance of leukemia cells and limited medical resources. Challenges arise from the subjectivity of evaluations and constraints in datasets, impacting the accuracy of classification. Existing methods face difficulties in achieving precise localization and building robust classification models due to the complexities in morphology and variations in subtypes, leading to challenges in accurate classification. This research proposes the FocusAugMix, a novel data augmentation method based on superpixels, which integrates Gradient-weighted Class Activation Mapping (Grad-CAM), Multi-Head Attention, and SaliencyMix to improve classification performance, especially in situations with limited datasets. The dynamic selection of superpixel contour images for each images allows this method to achieve a peak accuracy of 99.07 %, surpassing the previous method. Integrating Multi-Head Attention and Grad-CAM improves the accuracy and effectiveness of class representation in data augmentation methods for medical diagnosis.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"26 ","pages":"Article 200512"},"PeriodicalIF":0.0,"publicationDate":"2025-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143739146","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Explainable AI based LightGBM prediction model to predict default borrower in social lending platform","authors":"Li-Hua Li ,&nbsp;Alok Kumar Sharma ,&nbsp;Sheng-Tzong Cheng","doi":"10.1016/j.iswa.2025.200514","DOIUrl":"10.1016/j.iswa.2025.200514","url":null,"abstract":"<div><div>This paper proposes an explainable AI (XAI)-based prediction model utilizing the LightGBM algorithm to predict the likelihood of borrower default on a social lending platform. The dataset used in this study was obtained from Lending Club and consisted of various borrower characteristics and loan features. The proposed model not only provides high accuracy (0.87) in predicting defaulted borrowers, but also offers an explanation of the factors that contribute to the prediction. The model interpretability is facilitated through LIME and SHAP values, where SHAP values provide insights into the feature importance for the prediction. The outcome shows that the proposed model outperforms traditional approaches and delivers valuable insights for lending decision-making. The proposed model can be useful for lenders and regulators in the lending industry to improve decision-making processes and mitigating risk. Moreover, the XAI approach enables transparency and accountability in the decision-making process, making it more understandable and trustworthy for stakeholders.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"26 ","pages":"Article 200514"},"PeriodicalIF":0.0,"publicationDate":"2025-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143777643","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep learning-based CAD diagnosis using CNNs","authors":"Mohsen Amir Afzali,&nbsp;Hossein Ghaffarian","doi":"10.1016/j.iswa.2025.200507","DOIUrl":"10.1016/j.iswa.2025.200507","url":null,"abstract":"<div><div>Coronary Artery Disease (CAD) remains a significant global health concern, necessitating accurate diagnostic methods. In this study, we propose a deep learning solution for CAD diagnosis, driven by the limitations of traditional Machine Learning (ML) techniques in effectively handling numerical data. To address this, we focus exclusively on numerical features and employ essential preprocessing steps, including converting nominal features to numerical representations, normalizing numeric values, and balancing the dataset. Subsequently, we evaluate three deep learning classifiers—Convolutional Neural Network (CNN), Artificial Neural Network (ANN), and Long Short-Term Memory (LSTM)—to achieve improved diagnostic accuracy. Our evaluation of the proposed methods using real data demonstrates the superiority of deep learning techniques compared to other common classifiers, such as Random Forests, Bagging, Decision Trees, and Support Vector Machines (SVM). CNNs excel in feature extraction, capturing intricate patterns associated with CAD. Although ANNs and LSTMs are valuable, they do not match the discriminative power of CNNs in this context. In summary, our study underscores the pivotal role of CNNs in CAD diagnosis, achieving a highest accuracy of 98.64 %, representing a notable improvement compared to the best results reported in previous studies. This research not only advances the scientific understanding of CAD diagnostics but also has the potential to significantly enhance clinical practice by providing more accurate and timely diagnoses, ultimately improving patient outcomes and reducing healthcare costs.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"26 ","pages":"Article 200507"},"PeriodicalIF":0.0,"publicationDate":"2025-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143748711","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Integrated framework of fragment-based method and generative model for lead drug molecules discovery","authors":"Uche A.K. Chude-Okonkwo,&nbsp;Odifentse Lehasa","doi":"10.1016/j.iswa.2025.200508","DOIUrl":"10.1016/j.iswa.2025.200508","url":null,"abstract":"<div><div>Generative models have proven valuable in generating novel lead molecules with drug-like properties. However, beyond generating drug-like molecules, the generative model should also be able to create drug molecules with structural properties and pharmacophores to modulate a specific disease. The molecular generation process should also address the multi-objective optimization challenge of producing molecules with the desired efficacy and minimal side effects. This may entail the generation of a diverse pool of molecules with the desired structural properties and pharmacophore, which would offer diverse options and paths to developing potential new drug candidates by prioritizing molecules that balance the desired properties that can cater to the needs of different individuals. Achieving this requires a generative model learning a large dataset of molecular instances with the desired chemical/structural properties. However, large sets of drug molecules are not readily available for many diseases as there are few known drug molecular instances for treating any disease. To address this challenge, this paper presents an <em>in silico</em> molecular generative framework aided by fragment-based molecules’ synthesis for generating a pool of lead molecular instances possessing structural properties and pharmacophores to treat a disease of interest. The operation of the framework is explored using Hypertension as the disease of interest and beta-blocker as the reference hypertension drug to be generated. We generated over 123 beta-blocker-like molecules and further virtual-screened them for drug-likeness, docking probability, scaffold diversity, electrostatic complementarity, and synthesis accessibility to arrive at the final lead beta-blocker-like molecules.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"26 ","pages":"Article 200508"},"PeriodicalIF":0.0,"publicationDate":"2025-03-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143715269","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep learning for efficient high-resolution image processing: A systematic review","authors":"Albert Dede ,&nbsp;Henry Nunoo-Mensah ,&nbsp;Eric Tutu Tchao ,&nbsp;Andrew Selasi Agbemenu ,&nbsp;Prince Ebenezer Adjei ,&nbsp;Francisca Adoma Acheampong ,&nbsp;Jerry John Kponyo","doi":"10.1016/j.iswa.2025.200505","DOIUrl":"10.1016/j.iswa.2025.200505","url":null,"abstract":"<div><div>High-resolution images are increasingly used in fields such as remote sensing, medical imaging, and agriculture, but they present significant computational challenges when processed with deep learning models. This paper provides a systematic review of deep learning techniques developed to improve the efficiency of high-resolution image processing. We investigate techniques like lightweight neural networks, vision transformers adapted for high-resolution inputs, and models using frequency-domain inputs based on 96 studies from 2018 to 2023. These techniques have many applications, from environmental monitoring and urban planning to disease diagnosis. We emphasize the advancements in efficient high-resolution deep learning models, discussing their performance in terms of accuracy, speed, and resource requirements. Key challenges, including the trade-off between processing efficiency and model accuracy, are analysed, and potential future research directions are proposed to address these issues.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"26 ","pages":"Article 200505"},"PeriodicalIF":0.0,"publicationDate":"2025-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143705634","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Leveraging social networks as an optimization approach","authors":"Hamed Ghadirian ,&nbsp;Seyed Jalaleddin Mousavirad","doi":"10.1016/j.iswa.2025.200506","DOIUrl":"10.1016/j.iswa.2025.200506","url":null,"abstract":"<div><div>Metaheuristic algorithms have become powerful tools for solving complex optimization problems. Consensus-based optimization (CBO), inspired by social interactions, models a network where agents adjust their positions by learning from their neighbors. While effective, CBO relies on a fixed network structure, limiting its adaptability. To overcome this, we propose the Human Generation (HG) algorithm, which extends CBO by incorporating a two-layer influence mechanism. The first layer mimics kinship-based learning, ensuring local refinement, while the second layer models elite-following behavior, enabling efficient global exploration. This structured adaptation enhances both convergence speed and solution accuracy. We evaluate HG across unimodal, multimodal, and complex optimization problems, as well as a real-world image thresholding application. Experimental results demonstrate that HG consistently outperforms CBO and other state-of-the-art algorithms, making it a robust optimization approach.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"26 ","pages":"Article 200506"},"PeriodicalIF":0.0,"publicationDate":"2025-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143704010","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An effective approach for early fuel leakage detection with enhanced explainability","authors":"Ruimin Chu ,&nbsp;Li Chik ,&nbsp;Yiliao Song ,&nbsp;Jeffrey Chan ,&nbsp;Xiaodong Li","doi":"10.1016/j.iswa.2025.200504","DOIUrl":"10.1016/j.iswa.2025.200504","url":null,"abstract":"<div><div>Leakage detection at service stations with underground storage tanks containing hazardous products, such as fuel, is a critical task. Early detection is important to halt the spread of leaks, which can pose significant economic and ecological impacts on the surrounding community. Existing fuel leakage detection methods typically rely on statistical analysis of low-granularity inventory data, leading to delayed detection. Moreover, explainability, a crucial factor for practitioners to validate detection outcomes, remains unexplored in this domain. To address these limitations, we propose an <strong>EX</strong>plainable <strong>F</strong>uel <strong>L</strong>eakage <strong>D</strong>etection approach called EXFLD, which performs online fuel leakage detection and provides intuitive explanations for detection validation. EXFLD incorporates a high-performance deep learning model for accurate online fuel leakage detection and an inherently interpretable model to generate intuitive textual explanations to assist practitioners in result validation. Unlike existing explainable artificial intelligence methods that often use deep learning models which can be hard to interpret, EXFLD is a human-centric system designed to provide clear and understandable insights to support decision-making. Through case studies, we demonstrate that EXFLD can provide intuitive and meaningful textual explanations that humans can easily understand. Additionally, we show that incorporating semantic constraints during training in the ANFIS model enhances the semantic interpretability of these explanations by improving the coverage and distinguishability of membership functions. Experimental evaluations using a dataset collected from real-world sites in Australia, encompassing 167 tank instances, demonstrate that EXFLD achieves competitive performance compared to baseline methods, with an F2-score of 0.7969. This dual focus on accuracy and human-centric explainability marks a significant advancement in fuel leakage detection, potentially facilitating broader adoption.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"26 ","pages":"Article 200504"},"PeriodicalIF":0.0,"publicationDate":"2025-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143680449","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A systematic literature review of quantum object detection and recognition: research trend, datasets, topics and methods","authors":"Ifran Lindu Mahargya,&nbsp;Guruh Fajar Shidik,&nbsp;Affandy,&nbsp;Pujiono,&nbsp;Supriadi Rustad","doi":"10.1016/j.iswa.2025.200499","DOIUrl":"10.1016/j.iswa.2025.200499","url":null,"abstract":"<div><div>Quantum computing is a computational process that utilizes quantum mechanics features, namely superposition, interference, and entanglement, in information processing, allowing computation to run in parallel. The advantage of quantum computing is that it solves complex problems whereas classical computing is impossible because it requires expensive computing costs. Object detection and recognition is a task of computer vision, where research in this field aims to improve the ability of computer algorithms to produce interpretations of visual information. Humans easily analyze and describe the visual information received. However, unlike computer systems, they must learn and explore using machine learning from the visual information received to provide correct interpretations of visual information. This paper presents a systematic review of papers published from 2012 to 2024 to answer how far quantum object detection and recognition research has been conducted. The methodology of this review follows a systematic literature review such as the method proposed by Kitchenham et al. The selected primary studies amounted to 29 papers from four source digital libraries. The application of quantum algorithms is more often used to improve the performance of classical computing. The quantum model category consists of 3 types, namely pure quantum, hybrid classical-quantum, and quantum-inspired ML. Hybrid classical-quantum is the most discussed model and Quantum Convolutional Neural Network is the most frequently discussed algorithm or model in image classification from 2012 to 2024. Quantum algorithms show good results and can improve the performance of classical algorithms, although currently, the ability of quantum computing is not fully optimal because the development of quantum computers is still in the noisy intermediate-scale quantum era. However, with the current limited quantum computing capabilities, it can already outperform the capabilities of classical computing. Based on this, studies on quantum object detection and recognition need to be carried out so that when the full potential of quantum computing can be utilized, the user's capacity is competent.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"26 ","pages":"Article 200499"},"PeriodicalIF":0.0,"publicationDate":"2025-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143637305","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Learning how to transfer: A lifelong domain knowledge distillation framework for continual MRC","authors":"Songze Li ,&nbsp;Zhijing Wu ,&nbsp;Runmin Cao ,&nbsp;Xiaohan Zhang ,&nbsp;Yifan Wang ,&nbsp;Hua Xu ,&nbsp;Kai Gao","doi":"10.1016/j.iswa.2025.200497","DOIUrl":"10.1016/j.iswa.2025.200497","url":null,"abstract":"<div><div>Machine Reading Comprehension (MRC) has attracted wide attention in recent years. It can reflect how well a machine understands human language. Benefitting from the increasing large-scale benchmark and pre-trained language models, a lot of MRC models have achieved remarkable success and even exceeded human performance. However, real-world MRC systems need incrementally learn from a continuous data stream across time without accessing the previously seen data, called Continual MRC system. It is a great challenge to learn a new domain incrementally without catastrophically forgetting previous knowledge. In this paper, MK-MRC (an extension of MA-MRC), a continual MRC framework with uncertainty-aware fixed <strong>M</strong>emory and lifelong domain <strong>K</strong>nowledge distillation, is proposed. MK-MRC is a memory replaying based method, in which a fixed-size memory buffer stores a small number of samples in previous domain data along with an uncertainty-aware updating strategy when new domain data arrives. For incremental learning, MK-MRC fully uses the domain adaptation and transfer relationship between memory and new domain data through several domain knowledge distillation strategies.</div><div>Compared with MA-MRC, MK-MRC additionally introduces more strategies to strengthen the ability of continual learning, such as data augmentation and special task-related knowledge distillation. Experimental results show that MK-MRC yields consistent improvement compared with strong baselines and has a substantial incremental learning ability without catastrophically forgetting under four continual span-extractive and multiple-choice MRC settings.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"26 ","pages":"Article 200497"},"PeriodicalIF":0.0,"publicationDate":"2025-03-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143619117","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}