Teaching Learning Based Optimization (TLBO) algorithm, introduced in 2011 is widely used in optimization problems across various domains. It is a powerful tool that is capable of solving complex, multidimensional, linear, and nonlinear problems. MapReduce is a distributed programming model developed by Google. It is widely used for processing large datasets in parallel way. This paper proposes the use of the MapReduce programming paradigm for the implementation of the TLBO algorithm on distributed systems, creating a novel approach known as MapReduce Teaching Learning Based Optimization (MRTLBO). The proposed MRTLBO algorithm is tested on Congress of Evolutionary Computations (CEC)-2013 Large-Scale Global Optimization Benchmark Problems dataset, and its performance is compared with sequential TLBO algorithm on the same dataset. The experimental output exhibits that the MRTLBO algorithm is effective in working with high-dimensional problems, and it outperforms the sequential TLBO algorithm with respect to the final result, and speedup. Overall, the proposed MRTLBO algorithm gives a scalable and effective optimization strategy for working on optimization problems in distributed systems.
{"title":"MapReduce teaching learning based optimization algorithm for solving CEC-2013 LSGO benchmark Testsuit","authors":"A.J. Umbarkar , P.M. Sheth , Wei-Chiang Hong , S.M. Jagdeo","doi":"10.1016/j.iswa.2024.200460","DOIUrl":"10.1016/j.iswa.2024.200460","url":null,"abstract":"<div><div>Teaching Learning Based Optimization (TLBO) algorithm, introduced in 2011 is widely used in optimization problems across various domains. It is a powerful tool that is capable of solving complex, multidimensional, linear, and nonlinear problems. MapReduce is a distributed programming model developed by Google. It is widely used for processing large datasets in parallel way. This paper proposes the use of the MapReduce programming paradigm for the implementation of the TLBO algorithm on distributed systems, creating a novel approach known as MapReduce Teaching Learning Based Optimization (MRTLBO). The proposed MRTLBO algorithm is tested on Congress of Evolutionary Computations (CEC)-2013 Large-Scale Global Optimization Benchmark Problems dataset, and its performance is compared with sequential TLBO algorithm on the same dataset. The experimental output exhibits that the MRTLBO algorithm is effective in working with high-dimensional problems, and it outperforms the sequential TLBO algorithm with respect to the final result, and speedup. Overall, the proposed MRTLBO algorithm gives a scalable and effective optimization strategy for working on optimization problems in distributed systems.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"24 ","pages":"Article 200460"},"PeriodicalIF":0.0,"publicationDate":"2024-11-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142706536","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}
Pub Date : 2024-11-12DOI: 10.1016/j.iswa.2024.200459
Yong Wang, Jianfeng Zeng, Pengfei Du, Huachao Xu
The gear decision logic of automatic transmission directly affects the vehicle's dynamic, fuel economic, and comfort performance. This study employs data mining techniques to address the issues of low adaptability and low recognition rate in the intelligent gear decision of vehicle automatic transmission systems. The research further proposes the utilization of Kalman filter, Hidden Markov Models, and Long Short-Term Memory networks for condition feature recognition and time series classification. Subsequently, dynamic programming algorithms are employed to optimize intelligent gear decisions. Combining driver intent and driving environment, an intelligent gear decision method is formulated. The results indicate that, during a 430 s driving segment, the intelligent gear decision method consumes only 464 mL of fuel, closely resembling the economic strategy's 457 mL, with a gear shift frequency of 53, significantly better than the 79 shifts in the economic strategy. Moreover, the error rate for slope condition recognition is only 0.062 %. In a 200 s coupled condition, the intelligent gear decision results in fuel consumption of 207 mL, approximating the actual vehicle's 219 mL, while power-shifting consumes 316 mL, and economic shifting only 202mL. This study not only improves the accuracy of gear decisions but also effectively enhances vehicle operational efficiency, providing valuable insights for future automatic transmission systems with significant practical value.
{"title":"Intelligent gear decision method for vehicle automatic transmission system based on data mining","authors":"Yong Wang, Jianfeng Zeng, Pengfei Du, Huachao Xu","doi":"10.1016/j.iswa.2024.200459","DOIUrl":"10.1016/j.iswa.2024.200459","url":null,"abstract":"<div><div>The gear decision logic of automatic transmission directly affects the vehicle's dynamic, fuel economic, and comfort performance. This study employs data mining techniques to address the issues of low adaptability and low recognition rate in the intelligent gear decision of vehicle automatic transmission systems. The research further proposes the utilization of Kalman filter, Hidden Markov Models, and Long Short-Term Memory networks for condition feature recognition and time series classification. Subsequently, dynamic programming algorithms are employed to optimize intelligent gear decisions. Combining driver intent and driving environment, an intelligent gear decision method is formulated. The results indicate that, during a 430 s driving segment, the intelligent gear decision method consumes only 464 mL of fuel, closely resembling the economic strategy's 457 mL, with a gear shift frequency of 53, significantly better than the 79 shifts in the economic strategy. Moreover, the error rate for slope condition recognition is only 0.062 %. In a 200 s coupled condition, the intelligent gear decision results in fuel consumption of 207 mL, approximating the actual vehicle's 219 mL, while power-shifting consumes 316 mL, and economic shifting only 202mL. This study not only improves the accuracy of gear decisions but also effectively enhances vehicle operational efficiency, providing valuable insights for future automatic transmission systems with significant practical value.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"24 ","pages":"Article 200459"},"PeriodicalIF":0.0,"publicationDate":"2024-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142706627","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}
This paper presents a method to construct and implement an Events Knowledge Graph (EventsKG) for security-related open-source intelligence gathering, focusing on event exploration for situation monitoring in India. The EventsKG is designed to process news articles, extract events of national security significance, and represent them in a consistent and intuitive manner. This method utilizes state-of-the-art natural language understanding techniques and the capabilities of graph databases to extract and organize events. A domain-specific ontology is created for effective storage and retrieval. In addition, we provide a user-friendly dashboard for querying and a complete visualization of events across India. The effectiveness of the EventsKG is assessed through a human evaluation of the information retrieval quality. Our approach contributes to rapid data availability and decision-making through a comprehensive understanding of events, including local events, from every part of India in just a few clicks. The system is evaluated against a manually annotated dataset and by involving human evaluators through a feedback survey, and it has shown good retrieval accuracy. The EventsKG can also be used for other applications such as threat intelligence, incident response, and situational awareness.
{"title":"Design and implementation of EventsKG for situational monitoring and security intelligence in India: An open-source intelligence gathering approach","authors":"Hashmy Hassan , Sudheep Elayidom , M.R. Irshad , Christophe Chesneau","doi":"10.1016/j.iswa.2024.200458","DOIUrl":"10.1016/j.iswa.2024.200458","url":null,"abstract":"<div><div>This paper presents a method to construct and implement an Events Knowledge Graph (EventsKG) for security-related open-source intelligence gathering, focusing on event exploration for situation monitoring in India. The EventsKG is designed to process news articles, extract events of national security significance, and represent them in a consistent and intuitive manner. This method utilizes state-of-the-art natural language understanding techniques and the capabilities of graph databases to extract and organize events. A domain-specific ontology is created for effective storage and retrieval. In addition, we provide a user-friendly dashboard for querying and a complete visualization of events across India. The effectiveness of the EventsKG is assessed through a human evaluation of the information retrieval quality. Our approach contributes to rapid data availability and decision-making through a comprehensive understanding of events, including local events, from every part of India in just a few clicks. The system is evaluated against a manually annotated dataset and by involving human evaluators through a feedback survey, and it has shown good retrieval accuracy. The EventsKG can also be used for other applications such as threat intelligence, incident response, and situational awareness.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"24 ","pages":"Article 200458"},"PeriodicalIF":0.0,"publicationDate":"2024-11-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142658208","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}
Pub Date : 2024-11-08DOI: 10.1016/j.iswa.2024.200456
Kamalakkannan Ravi, Jiann-Shiun Yuan
The rise of social networking sites has reshaped digital interactions, becoming fertile grounds for extremist ideologies, notably in the United States. Despite previous research, understanding and tackling online ideological extremism remains challenging. In this context, we conduct a systematic literature review to comprehensively analyze existing research and offer insights for both researchers and policymakers. Spanning from 2005 to 2023, our review includes 110 primary research articles across platforms like Twitter (X), Facebook, Reddit, TikTok, Telegram, and Parler. We observe a diverse array of methodologies, including natural language processing (NLP), machine learning (ML), deep learning (DL), graph-based methods, dictionary-based methods, and statistical approaches. Through synthesis, we aim to advance understanding and provide actionable recommendations for combating ideological extremism effectively on online social networking sites.
{"title":"Ideological orientation and extremism detection in online social networking sites: A systematic review","authors":"Kamalakkannan Ravi, Jiann-Shiun Yuan","doi":"10.1016/j.iswa.2024.200456","DOIUrl":"10.1016/j.iswa.2024.200456","url":null,"abstract":"<div><div>The rise of social networking sites has reshaped digital interactions, becoming fertile grounds for extremist ideologies, notably in the United States. Despite previous research, understanding and tackling online ideological extremism remains challenging. In this context, we conduct a systematic literature review to comprehensively analyze existing research and offer insights for both researchers and policymakers. Spanning from 2005 to 2023, our review includes 110 primary research articles across platforms like Twitter (X), Facebook, Reddit, TikTok, Telegram, and Parler. We observe a diverse array of methodologies, including natural language processing (NLP), machine learning (ML), deep learning (DL), graph-based methods, dictionary-based methods, and statistical approaches. Through synthesis, we aim to advance understanding and provide actionable recommendations for combating ideological extremism effectively on online social networking sites.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"24 ","pages":"Article 200456"},"PeriodicalIF":0.0,"publicationDate":"2024-11-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142658207","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}
Pub Date : 2024-10-31DOI: 10.1016/j.iswa.2024.200452
Peifang Liu , Jiang Guo , Fangqing Zhang , Ye Zou , Junjie Tang
In the ever-evolving landscape of power networks, the integration of diverse sources, including electric vehicles (EVs) and renewable energies like wind power, has gained prominence. With the rapid proliferation of plug-in electric vehicles (PEVs), their optimal utilization hinges on reconciling conflicting and adaptable targets, including facilitating vehicle-to-grid (V2 G) connectivity or harmonizing with the broader energy ecosystem. Simultaneously, the inexorable integration of wind resources into power networks underscores the critical need for multi-purpose planning to optimize production and reduce costs. This study tackles this multifaceted challenge, incorporating the presence of EVs and a probabilistic wind resource model. Addressing the complexity of the issue, we devise a multi-purpose method grounded in collective competition, effectively reducing computational complexity and creatively enhancing the model's performance with a Pareto front optimality point. To discern the ideal response, fuzzy theory is employed. The suggested pattern is rigorously tested on two well-established IEEE power networks (30- and 118-bus) in diverse scenarios featuring windmills and PEV producers, with outcomes showcasing the remarkable excellence of our multi-purpose framework in addressing this intricate issue while accommodating uncertainty.
{"title":"Multi-objective optimization of power networks integrating electric vehicles and wind energy","authors":"Peifang Liu , Jiang Guo , Fangqing Zhang , Ye Zou , Junjie Tang","doi":"10.1016/j.iswa.2024.200452","DOIUrl":"10.1016/j.iswa.2024.200452","url":null,"abstract":"<div><div>In the ever-evolving landscape of power networks, the integration of diverse sources, including electric vehicles (EVs) and renewable energies like wind power, has gained prominence. With the rapid proliferation of plug-in electric vehicles (PEVs), their optimal utilization hinges on reconciling conflicting and adaptable targets, including facilitating vehicle-to-grid (V2 G) connectivity or harmonizing with the broader energy ecosystem. Simultaneously, the inexorable integration of wind resources into power networks underscores the critical need for multi-purpose planning to optimize production and reduce costs. This study tackles this multifaceted challenge, incorporating the presence of EVs and a probabilistic wind resource model. Addressing the complexity of the issue, we devise a multi-purpose method grounded in collective competition, effectively reducing computational complexity and creatively enhancing the model's performance with a Pareto front optimality point. To discern the ideal response, fuzzy theory is employed. The suggested pattern is rigorously tested on two well-established IEEE power networks (30- and 118-bus) in diverse scenarios featuring windmills and PEV producers, with outcomes showcasing the remarkable excellence of our multi-purpose framework in addressing this intricate issue while accommodating uncertainty.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"24 ","pages":"Article 200452"},"PeriodicalIF":0.0,"publicationDate":"2024-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142658206","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}
Pub Date : 2024-10-29DOI: 10.1016/j.iswa.2024.200455
Charlotte Jamotton, Donatien Hainaut
This article explores the application of Variational AutoEncoders (VAEs) to insurance data. Previous research has demonstrated the successful implementation of generative models, especially VAEs, across various domains, such as image recognition, text classification, and recommender systems. However, their application to insurance data, particularly to heterogeneous insurance portfolios with mixed continuous and discrete attributes, remains unexplored. This study introduces novel insights into utilising VAEs for unsupervised learning tasks in actuarial science, including dimension reduction and synthetic data generation. We propose a VAE model with a quantile transformation for continuous (latent) variables, a reconstruction loss that combines categorical cross-entropy and mean squared error, and a KL divergence-based regularisation term. Our VAE model’s architecture circumvents the need to pre-train and fine-tune a neural network to encode categorical variables into -dimensional representative vectors within a continuous vector space of dimension . We assess our VAE’s ability to reconstruct complex insurance data and generate synthetic insurance policies using a motor portfolio. Our experimental results and analysis highlight the potential of VAEs in addressing challenges related to data availability in the insurance industry.
{"title":"Variational AutoEncoder for synthetic insurance data","authors":"Charlotte Jamotton, Donatien Hainaut","doi":"10.1016/j.iswa.2024.200455","DOIUrl":"10.1016/j.iswa.2024.200455","url":null,"abstract":"<div><div>This article explores the application of Variational AutoEncoders (VAEs) to insurance data. Previous research has demonstrated the successful implementation of generative models, especially VAEs, across various domains, such as image recognition, text classification, and recommender systems. However, their application to insurance data, particularly to heterogeneous insurance portfolios with mixed continuous and discrete attributes, remains unexplored. This study introduces novel insights into utilising VAEs for unsupervised learning tasks in actuarial science, including dimension reduction and synthetic data generation. We propose a VAE model with a quantile transformation for continuous (latent) variables, a reconstruction loss that combines categorical cross-entropy and mean squared error, and a KL divergence-based regularisation term. Our VAE model’s architecture circumvents the need to pre-train and fine-tune a neural network to encode categorical variables into <span><math><mi>n</mi></math></span>-dimensional representative vectors within a continuous vector space of dimension <span><math><msup><mrow><mi>R</mi></mrow><mrow><mi>n</mi></mrow></msup></math></span>. We assess our VAE’s ability to reconstruct complex insurance data and generate synthetic insurance policies using a motor portfolio. Our experimental results and analysis highlight the potential of VAEs in addressing challenges related to data availability in the insurance industry.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"24 ","pages":"Article 200455"},"PeriodicalIF":0.0,"publicationDate":"2024-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142592550","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}
Pub Date : 2024-10-28DOI: 10.1016/j.iswa.2024.200454
Cathaoir Agnew , Eoin M. Grua , Pepijn Van de Ven , Patrick Denny , Ciarán Eising , Anthony Scanlan
Annotating datasets for fully supervised instance segmentation tasks can be arduous and time-consuming, requiring a significant effort and cost investment. Producing bounding box annotations instead constitutes a significant reduction in this investment, but bounding box annotated data alone are not suitable for instance segmentation. This work utilizes ground truth bounding boxes to define coarsely annotated polygon masks, which we refer to as weak annotations, on which the models are pre-trained. We investigate the effect of pretraining on data with weak annotations and further fine-tuning on data with strong annotations, that is, finely annotated polygon masks for instance segmentation. The COCO 2017 detection dataset along with 3 model architectures, SOLOv2, Mask-RCNN, and Mask2former, were used to conduct experiments investigating the effect of pretraining on weak annotations. The Cityscapes and Pascal VOC 2012 datasets were used to validate this approach. The empirical results suggest two key outcomes from this investigation. Firstly, a sequential approach to annotating large-scale instance segmentation datasets would be beneficial, enabling higher-performance models in faster timeframes. This is accomplished by first labeling bounding boxes on your data followed by polygon masks. Secondly, it is possible to leverage object detection datasets for pretraining instance segmentation models while maintaining competitive results in the downstream task. This is reflected with 97.5%, 100.4% & 101.3% of the fully supervised performance being achieved with just 1%, 5% & 10% of the instance segmentation annotations of the COCO training dataset being utilized for the best performing model, Mask2former with a Swin-L backbone.
{"title":"Pretraining instance segmentation models with bounding box annotations","authors":"Cathaoir Agnew , Eoin M. Grua , Pepijn Van de Ven , Patrick Denny , Ciarán Eising , Anthony Scanlan","doi":"10.1016/j.iswa.2024.200454","DOIUrl":"10.1016/j.iswa.2024.200454","url":null,"abstract":"<div><div>Annotating datasets for fully supervised instance segmentation tasks can be arduous and time-consuming, requiring a significant effort and cost investment. Producing bounding box annotations instead constitutes a significant reduction in this investment, but bounding box annotated data alone are not suitable for instance segmentation. This work utilizes ground truth bounding boxes to define coarsely annotated polygon masks, which we refer to as weak annotations, on which the models are pre-trained. We investigate the effect of pretraining on data with weak annotations and further fine-tuning on data with strong annotations, that is, finely annotated polygon masks for instance segmentation. The COCO 2017 detection dataset along with 3 model architectures, SOLOv2, Mask-RCNN, and Mask2former, were used to conduct experiments investigating the effect of pretraining on weak annotations. The Cityscapes and Pascal VOC 2012 datasets were used to validate this approach. The empirical results suggest two key outcomes from this investigation. Firstly, a sequential approach to annotating large-scale instance segmentation datasets would be beneficial, enabling higher-performance models in faster timeframes. This is accomplished by first labeling bounding boxes on your data followed by polygon masks. Secondly, it is possible to leverage object detection datasets for pretraining instance segmentation models while maintaining competitive results in the downstream task. This is reflected with 97.5%, 100.4% & 101.3% of the fully supervised performance being achieved with just 1%, 5% & 10% of the instance segmentation annotations of the COCO training dataset being utilized for the best performing model, Mask2former with a Swin-L backbone.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"24 ","pages":"Article 200454"},"PeriodicalIF":0.0,"publicationDate":"2024-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142572831","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}
Cardiomyopathy is a serious condition that can result in heart failure, sudden cardiac death, malignant arrhythmias, and thromboembolism. It is a significant contributor to morbidity and mortality globally. The initial finding of cardiomegaly on radiological imaging may signal a deterioration of a known heart condition, an unknown heart disease, or a heart complication related to another illness. Further cardiological evaluation is needed to confirm the diagnosis and determine appropriate treatment. A chest radiograph (X-ray) is the main imaging method used to identify cardiomegaly when the heart is enlarged. A prompt and accurate diagnosis is essential to help healthcare providers determine the most appropriate treatment options before the condition worsens. This study aims to utilize convolutional neural networks and transfer learning techniques, specifically Inception, DenseNet-169, and ResNet-50, to classify cardiomegaly from chest X-ray images automatically. The utilization of block-matching and 3D filtering (BM3D) techniques aimed at enhancing image edge retention, decreasing noise, and utilizing contrast limited adaptive histogram equalization (CLAHE) to enhance contrast in low-intensity images. Gradient-weighted Class Activation Mapping (GradCAM) was used to visualize the significant activation regions contributing to the model's decision. After evaluating all the models, the ResNet-50 model showed outstanding performance. It achieved perfect accuracy of 100 % in both training, and validation, and an impressive 99.8 % accuracy in testing. Additionally, it displayed complete 100 % precision, recall, and F1-score. These findings demonstrate that ResNet-50 surpassed all other models in the study. As a result, the impressive performance of the ResNet-50 model suggests that it could be a valuable tool in improving the efficiency and accuracy of cardiomyopathy diagnosis, ultimately leading to better patient outcomes.
心肌病是一种可导致心力衰竭、心脏性猝死、恶性心律失常和血栓栓塞的严重疾病。它是全球发病率和死亡率的重要因素。最初在放射成像中发现心脏肿大可能预示着已知心脏疾病、未知心脏疾病或与其他疾病相关的心脏并发症的恶化。需要进一步的心脏病学评估来确诊并确定适当的治疗方法。当心脏增大时，胸片（X 光）是确定心脏肥大的主要影像学方法。及时准确的诊断对于帮助医疗人员在病情恶化前确定最合适的治疗方案至关重要。本研究旨在利用卷积神经网络和迁移学习技术（特别是 Inception、DenseNet-169 和 ResNet-50）自动对胸部 X 光图像中的心脏肿大进行分类。利用块匹配和三维滤波（BM3D）技术，旨在增强图像边缘保留、降低噪声，并利用对比度受限自适应直方图均衡（CLAHE）增强低强度图像的对比度。梯度加权类激活图谱（GradCAM）用于可视化对模型决策有贡献的重要激活区域。在对所有模型进行评估后，ResNet-50 模型表现突出。它在训练和验证中都达到了 100% 的完美准确率，在测试中的准确率也达到了令人印象深刻的 99.8%。此外，它的精确度、召回率和 F1 分数都达到了 100%。这些结果表明，ResNet-50 超越了研究中的所有其他模型。因此，ResNet-50 模型令人印象深刻的表现表明，它可以成为提高心肌病诊断效率和准确性的重要工具，最终为患者带来更好的治疗效果。
{"title":"Early-stage cardiomegaly detection and classification from X-ray images using convolutional neural networks and transfer learning","authors":"Aleka Melese Ayalew , Belay Enyew , Yohannes Agegnehu Bezabh , Biniyam Mulugeta Abuhayi , Girma Sisay Negashe","doi":"10.1016/j.iswa.2024.200453","DOIUrl":"10.1016/j.iswa.2024.200453","url":null,"abstract":"<div><div>Cardiomyopathy is a serious condition that can result in heart failure, sudden cardiac death, malignant arrhythmias, and thromboembolism. It is a significant contributor to morbidity and mortality globally. The initial finding of cardiomegaly on radiological imaging may signal a deterioration of a known heart condition, an unknown heart disease, or a heart complication related to another illness. Further cardiological evaluation is needed to confirm the diagnosis and determine appropriate treatment. A chest radiograph (X-ray) is the main imaging method used to identify cardiomegaly when the heart is enlarged. A prompt and accurate diagnosis is essential to help healthcare providers determine the most appropriate treatment options before the condition worsens. This study aims to utilize convolutional neural networks and transfer learning techniques, specifically Inception, DenseNet-169, and ResNet-50, to classify cardiomegaly from chest X-ray images automatically. The utilization of block-matching and 3D filtering (BM3D) techniques aimed at enhancing image edge retention, decreasing noise, and utilizing contrast limited adaptive histogram equalization (CLAHE) to enhance contrast in low-intensity images. Gradient-weighted Class Activation Mapping (GradCAM) was used to visualize the significant activation regions contributing to the model's decision. After evaluating all the models, the ResNet-50 model showed outstanding performance. It achieved perfect accuracy of 100 % in both training, and validation, and an impressive 99.8 % accuracy in testing. Additionally, it displayed complete 100 % precision, recall, and F1-score. These findings demonstrate that ResNet-50 surpassed all other models in the study. As a result, the impressive performance of the ResNet-50 model suggests that it could be a valuable tool in improving the efficiency and accuracy of cardiomyopathy diagnosis, ultimately leading to better patient outcomes.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"24 ","pages":"Article 200453"},"PeriodicalIF":0.0,"publicationDate":"2024-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142525804","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}
Metaheuristics can benefit from analyzing patterns and regularities in data to perform more effective searches in the solution space. In line with the emerging trend in the optimization literature, this study introduces the Reinforcement-learning-based Alpha-List Iterated Greedy (RAIG) algorithm to contribute to the advances in machine learning-based optimization, notably for solving combinatorial problems. RAIG uses an N-List mechanism for solution initialization and its solution improvement procedure is enhanced by Reinforcement Learning and an Alpha-List mechanism for more effective searches. A classic engineering optimization problem, the Permutation Flowshop Scheduling Problem (PFSP), is considered for numerical experiments to evaluate RAIG's performance. Highly competitive solutions to the classic scheduling problem are identified, with up to 9% improvement compared to the baseline, when solving large-size instances. Experimental results also show that the RAIG algorithm performs more robustly than the baseline algorithm. Statistical tests confirm that RAIG is superior and hence can be introduced as a strong benchmark for future studies.
元启发式算法可以从分析数据中的模式和规律性中获益，从而在解空间中进行更有效的搜索。根据优化文献中的新兴趋势，本研究引入了基于强化学习的阿尔法列表迭代贪婪算法（RAIG），为基于机器学习的优化（尤其是解决组合问题）的进步做出贡献。RAIG 采用 N 列表机制进行求解初始化，其求解改进程序通过强化学习和 Alpha 列表机制得到增强，从而实现更有效的搜索。为了评估 RAIG 的性能，我们在数值实验中考虑了一个经典的工程优化问题，即 Permutation Flowshop Scheduling Problem (PFSP)。与基线相比，在求解大型实例时，RAIG 的性能提高了 9%。实验结果还表明，RAIG 算法比基准算法更稳健。统计测试证实了 RAIG 算法的优越性，因此可以作为未来研究的有力基准。
{"title":"Reinforcement learning-based alpha-list iterated greedy for production scheduling","authors":"Kuo-Ching Ying , Pourya Pourhejazy , Shih-Han Cheng","doi":"10.1016/j.iswa.2024.200451","DOIUrl":"10.1016/j.iswa.2024.200451","url":null,"abstract":"<div><div>Metaheuristics can benefit from analyzing patterns and regularities in data to perform more effective searches in the solution space. In line with the emerging trend in the optimization literature, this study introduces the Reinforcement-learning-based Alpha-List Iterated Greedy (RAIG) algorithm to contribute to the advances in machine learning-based optimization, notably for solving combinatorial problems. RAIG uses an <em>N</em>-List mechanism for solution initialization and its solution improvement procedure is enhanced by Reinforcement Learning and an Alpha-List mechanism for more effective searches. A classic engineering optimization problem, the Permutation Flowshop Scheduling Problem (PFSP), is considered for numerical experiments to evaluate RAIG's performance. Highly competitive solutions to the classic scheduling problem are identified, with up to 9% improvement compared to the baseline, when solving large-size instances. Experimental results also show that the RAIG algorithm performs more robustly than the baseline algorithm. Statistical tests confirm that RAIG is superior and hence can be introduced as a strong benchmark for future studies.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"24 ","pages":"Article 200451"},"PeriodicalIF":0.0,"publicationDate":"2024-10-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142441591","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}
In this paper, a machine learning approach is proposed to predict the next day's stock prices. The methodology involves comprehensive data collection and feature generation, followed by predictions utilizing Multi-Layer Perceptron (MLP) networks. We selected 5,283 records of daily historical data, including open prices, close prices, highest prices, lowest prices, and trading volumes from four well-known stocks in the FTSE 100 index. A novel set of engineered and derivative indices is extracted from the original time series to enhance prediction accuracy. Two Multi-Layer Perceptron (MLP) are proposed to predict the next day's stock prices using the engineered discrete and continuous indices. The case study uses the daily historical time series of stock prices between January 1, 2000, and December 31, 2020. The proposed machine learning approach presents suitable applicability and accuracy, respectively.
{"title":"A multi-stage machine learning approach for stock price prediction: Engineered and derivative indices","authors":"Shaghayegh Abolmakarem , Farshid Abdi , Kaveh Khalili-Damghani , Hosein Didehkhani","doi":"10.1016/j.iswa.2024.200449","DOIUrl":"10.1016/j.iswa.2024.200449","url":null,"abstract":"<div><div>In this paper, a machine learning approach is proposed to predict the next day's stock prices. The methodology involves comprehensive data collection and feature generation, followed by predictions utilizing Multi-Layer Perceptron (MLP) networks. We selected 5,283 records of daily historical data, including open prices, close prices, highest prices, lowest prices, and trading volumes from four well-known stocks in the FTSE 100 index. A novel set of engineered and derivative indices is extracted from the original time series to enhance prediction accuracy. Two Multi-Layer Perceptron (MLP) are proposed to predict the next day's stock prices using the engineered discrete and continuous indices. The case study uses the daily historical time series of stock prices between January 1, 2000, and December 31, 2020. The proposed machine learning approach presents suitable applicability and accuracy, respectively.</div></div>","PeriodicalId":100684,"journal":{"name":"Intelligent Systems with Applications","volume":"24 ","pages":"Article 200449"},"PeriodicalIF":0.0,"publicationDate":"2024-10-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142425942","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}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号