Applied Soft Computing最新文献

{"title":"Stochastic consensus for uncertain multiple attribute group decision-making problem in belief distribution environment","authors":"Xianchao Dai ,&nbsp;Hao Li ,&nbsp;Ligang Zhou ,&nbsp;Qun Wu","doi":"10.1016/j.asoc.2024.112495","DOIUrl":"10.1016/j.asoc.2024.112495","url":null,"abstract":"<div><div>In the realm of uncertain multiple attribute group decision-making (MAGDM) problems, existing research often focuses on the development of consensus-enhancing algorithms grounded in optimization models. However, this paper takes a stochastic perspective, thoroughly considering the impact of uncertainty on decision-making. And a novel stochastic method to model group consensus is introduced with the listed three components: (1) the concept of stochastic rank analysis based on stochastic belief distribution (BD) is given to measure the uncertainty degree in the original BD matrix, which is then used to assign weights to decision makers (DMs). (2) in uncertain environments, to ensure the effectiveness of consensus from a probabilistic perspective, the stochastic consensus index is proposed by taking both the advantages of the Jensen-Shannon (JS) distance and the hesitant distance between stochastic BDs. Then, the expected acceptable group consensus index is further provided to measure the consensus of original preferences among the group, and (3) finally, to deal with the issue of no consensus information, an optimization model is constructed aimed at achieving an acceptable consensus that can generate recommendation advice for DMs, facilitating the attainment of a consensus. The effectiveness of the proposed method is exemplified through two case studies: purchase of new energy vehicles (NEVs) and a postgraduate interview scenario. Furthermore, sensitivity analysis and comparative analysis are presented to better prove its advantages.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"168 ","pages":"Article 112495"},"PeriodicalIF":7.2,"publicationDate":"2024-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142745490","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Tourism forecast combination using weighting schemes with flow information among component models","authors":"Yi-Chung Hu","doi":"10.1016/j.asoc.2024.112498","DOIUrl":"10.1016/j.asoc.2024.112498","url":null,"abstract":"<div><div>Forecast combination is an effective way to improve the accuracy of tourism demand forecasting. The continuous development of forecast combination methods with high accuracy is inevitable to help tourism practitioners formulate more appropriate management strategies. This study investigated how tourism forecasting accuracy can be improved by treating combination forecasting as a multiple attribute decision making (MADM) problem. The proposed hybrid methods first yield single-model forecasts from grey models without considering the sample size and limiting the available data to satisfy any statistical properties. Given the effectiveness of PROMETHEE in MADM, which applies flows to gauge the intensity of the preference for one alternative over another, the flows among component models in a combination are then used to assess relative weights next. Finally, the flow-based weighting schemes are incorporated into the linear and nonlinear combinations of individual forecasts. After assessing the accuracy of the proposed methods with the inbound tourism demand in Taiwan, the results indicated that the proposed methods involving the integration of the flow-based weighting scheme into the Choquet fuzzy integral performed better than other benchmark forecast combination methods with different model combinations.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"168 ","pages":"Article 112498"},"PeriodicalIF":7.2,"publicationDate":"2024-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142723390","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Federated graph neural network for privacy-preserved supply chain data sharing","authors":"Xiaochuan Tang ,&nbsp;Yu Wang ,&nbsp;Xin Liu ,&nbsp;Xiaojun Yuan ,&nbsp;Chao Fan ,&nbsp;Yanmei Hu ,&nbsp;Qiang Miao","doi":"10.1016/j.asoc.2024.112475","DOIUrl":"10.1016/j.asoc.2024.112475","url":null,"abstract":"<div><div>Machine learning plays an increasingly important role in supply chain management. Due to privacy and security concerns, enterprises are reluctant to share their raw data, which leads to missing links in supply chains. To address privacy issue and promote data sharing in supply chain, we propose a new federated graph neural network named Isomorphic Federated Graph Neural Network (IFGNN) for supply chain data sharing. IFGNN consists of a server and multiple clients. The server is a lightweight parameter server with an efficient parameter updating algorithm. A client is assigned to each node in the supply chain network. Every supplier client is linked to its first-order neighbors, which means they have supply–demand relationship. The topology of the input network is identical to that of the supplier clients. Experimental results on a newly collected vehicle supply chain dataset show that the performance of IFGNN is close to its centralized counterpart. This work demonstrates that it is feasible to protect supply chain data privacy without a significant loss of prediction accuracy. Federated learning provides a new solution for promoting data sharing and collaborative machine learning in supply chain.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"168 ","pages":"Article 112475"},"PeriodicalIF":7.2,"publicationDate":"2024-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142723388","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An intelligent model of pulmonary emphysema detection using adaptive image segmentation and multi-dilated densenet with attention mechanism","authors":"Indira Linginani,&nbsp;Muddana A. Lakshmi","doi":"10.1016/j.asoc.2024.112483","DOIUrl":"10.1016/j.asoc.2024.112483","url":null,"abstract":"<div><div>Pulmonary emphysema is a significant factor in lung cancer and Chronic Obstructive Pulmonary Disease (COPD). Traditional pulmonary emphysema detection models may have difficulty in accurately detecting and diagnosing the severity of the disease. So, this work developed a novel pulmonary emphysema detection system with the help of deep learning frameworks. Originally, the significant images are accumulated from the benchmark sources, and fed into the Adaptive Trans-DenseUnet (ATDUnet)-based segmentation model. The ATDUnet model is highly effective in accurately segmenting pulmonary emphysema from the gathered images. Moreover, to enhance the segmentation process, the parameters are tuned in the ATDUnet using the Statistical Solution of the Osprey Optimization Algorithm (SSOOA). Subsequently, the segmented image is given to the pulmonary emphysema classification phase, where the Multi-Dilated DenseNet with Attention Mechanism (MDDNet-AM) is employed. By incorporating an attention mechanism, MDDNet-AM can focus on important features within the image for improved accuracy and efficiency in diagnosis. Finally, the developed model offered the pulmonary emphysema classified outcome. Then, the outcome of the developed model is compared against conventional pulmonary emphysema detection methods, and given the accuracy to be 93.10. Therefore, the result proved that the use of developed advanced technology in pulmonary emphysema detection has shown promising results in the field of respiratory health.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"168 ","pages":"Article 112483"},"PeriodicalIF":7.2,"publicationDate":"2024-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142745497","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Optimized hybrid XGBoost-CatBoost model for enhanced prediction of concrete strength and reliability analysis using Monte Carlo simulations","authors":"Tuan Nguyen-Sy","doi":"10.1016/j.asoc.2024.112490","DOIUrl":"10.1016/j.asoc.2024.112490","url":null,"abstract":"<div><div>Building on our previous work demonstrating the exceptional potential of the Extreme Gradient Boosting model (XGBoost) for predicting the uniaxial compressive strength of concrete, this study introduces several significant advancements. First, we develop a novel optimized hybrid model that synergistically combines XGBoost, CatBoost (one of the most advanced tree-boosting models), and the Optuna algorithm to achieve unprecedented prediction accuracy. Second, we apply this hybrid model in Monte Carlo simulations to conduct a pioneering reliability analysis of concrete strength, capturing the effects of input uncertainty. Third, we propose an innovative technique for estimating tree leaf values, which fundamentally improves prediction accuracy. Our optimized hybrid model delivers outstanding performance, as evidenced by a five-fold cross-validation showing a coefficient of determination (R²) of 0.953, a root mean squared error (RMSE) of 3.603 MPa, and a mean absolute error (MAE) of 2.261 MPa—metrics that surpass the best results reported in the existing literature. Additionally, our Monte Carlo simulations reveal a substantial error range of 10–20 MPa for a ±5 % variation in input features, underscoring the critical impact of input uncertainty on prediction reliability. Furthermore, our new leaf value estimation technique significantly outperforms traditional averaging methods, offering a transformative improvement in model accuracy. These findings are crucial for broadening the scope of machine learning applications in civil engineering and other engineering disciplines.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"167 ","pages":"Article 112490"},"PeriodicalIF":7.2,"publicationDate":"2024-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142703501","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Anomaly-based network intrusion detection using denoising autoencoder and Wasserstein GAN synthetic attacks","authors":"Mohammad Arafah ,&nbsp;Iain Phillips ,&nbsp;Asma Adnane ,&nbsp;Wael Hadi ,&nbsp;Mohammad Alauthman ,&nbsp;Abedal-Kareem Al-Banna","doi":"10.1016/j.asoc.2024.112455","DOIUrl":"10.1016/j.asoc.2024.112455","url":null,"abstract":"<div><div>Intrusion detection systems face challenges in handling high-dimensional, large-scale, and imbalanced network traffic data. This paper proposes a novel architecture combining a denoising autoencoder (AE) and a Wasserstein generative adversarial network (WGAN) to address these issues. The AE-WGAN model extracts high-representative features and generates realistic synthetic attacks, effectively resolving data imbalance and enhancing anomaly-based intrusion detection. Extensive experiments on NSL-KDD and CICIDS-2017 datasets, using both binary and multiclass classification scenarios with various classifier architectures, demonstrate the model’s superior performance. The proposed approach outperforms state-of-the-art models in accuracy, precision, recall, and F1 score, showing excellent generalization capabilities against unseen attacks. Time complexity analysis reveals computational efficiency while maintaining high-quality synthetic attack generation. This research contributes a robust, efficient, and adaptable framework for intrusion detection, capable of handling modern network traffic complexities and evolving cyber threats.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"168 ","pages":"Article 112455"},"PeriodicalIF":7.2,"publicationDate":"2024-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142723431","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Z-number-based three-way decision method with classification-based state determination for the evaluation of new energy enterprises","authors":"Xiaowan Jin ,&nbsp;Huchang Liao ,&nbsp;Zhiying Zhang","doi":"10.1016/j.asoc.2024.112489","DOIUrl":"10.1016/j.asoc.2024.112489","url":null,"abstract":"<div><div>New energy enterprises are important promoting factors for sustainable development of modern society. Limited budgets of governments require that new energy enterprises should be efficiently evaluated before they are founded. Existing evaluation methods ignored the hesitation of experts to some alternatives. Although the three-way decision method has been applied widely as a method to evaluate alternatives, the determination of the state set has not been deeply discussed. To solve these challenges, this paper proposes a Z-number-based three-way decision method with classification-based state determination, which can assign alternatives with hesitation to a boundary region for further consideration and compute the conditional probability with a classification-based method. First, since traditional fuzzy sets cannot ensure the reliability of decision information, an evaluation matrix based on Z-numbers is constructed. Second, a fuzzy best-worst method is applied to determine the weights of criteria. Third, the conditional probability is computed based on classification-based state sets that are obtained by a sorting method. An example regarding the evaluation and selection of new energy enterprises demonstrates the validity and stability of the proposed method. The comparison analysis shows that our proposed method can divide alternatives into different regions efficiently and is less affected by the variation of parameters.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"167 ","pages":"Article 112489"},"PeriodicalIF":7.2,"publicationDate":"2024-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142703502","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A fuzzy grouping-based memetic algorithm for multi-depot multi-UAV power pole inspection","authors":"Xiang-Ling Chen ,&nbsp;Ya-Hui Jia ,&nbsp;Xiao-Cheng Liao ,&nbsp;Wei-Neng Chen","doi":"10.1016/j.asoc.2024.112472","DOIUrl":"10.1016/j.asoc.2024.112472","url":null,"abstract":"<div><div>Power pole inspection is important to maintain the normal operation of electrical system. It usually requires to fly a fleet of unmanned aerial vehicles (UAVs) from multiple depots at the same time to jointly complete the inspection tasks distributed in a wide area, which is a challenging planning problem. In order to address this problem, this work first builds the model of the multi-depot multi-UAV power pole inspection problem with charging stations. After that, a fuzzy grouping-based memetic algorithm named FGATS is proposed to solve the problem. Specifically, a fuzzy grouping strategy is proposed to divide a large-scale problem into multiple small-scale subproblems in order to reduce the complexity of the problem. It continuously adjusts the grouping scheme to enhance the flexibility and effectiveness of the algorithm. Then, a hybrid algorithm combining genetic algorithm and tabu search is designed to jointly optimize the subproblems, ensuring an effective balance between global and local searches. After a certain number of iterations, the problem is re-divided and the populations are re-initialized by the proposed solution update strategy that learns and incorporates historical task-sequence knowledge. This strategy enhances the current optimization process by retaining useful information from previous iterations. Experiments on both artificial terrains and a real terrain verifies the effectiveness of FGATS, with the algorithm’s performance ranking first overall.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"168 ","pages":"Article 112472"},"PeriodicalIF":7.2,"publicationDate":"2024-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142745492","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Adaptive deep learning models for efficient multivariate anomaly detection in IoT infrastructures","authors":"Ali Jameel Hashim ,&nbsp;M.A. Balafar ,&nbsp;Jafar Tanha ,&nbsp;Aryaz Baradarani","doi":"10.1016/j.asoc.2024.112377","DOIUrl":"10.1016/j.asoc.2024.112377","url":null,"abstract":"<div><div>Adapting machine learning-based systems to dynamic environments poses significant challenges due to their diverse and rapidly changing nature. Traditional Deep Neural Network (DNN) algorithms often struggle to cope effectively with such variations. This paper presents a novel evolutionary algorithm named Double Evaluation Genetic Evolution (DEGE), specifically tailored to evolve DNNs within dynamic contexts. DEGE represents a pioneering approach in evolutionary computing, focusing on the adaptive evolution of DNN structures across generations. This adaptability plays a crucial role in enabling DNNs to seamlessly adjust to evolving environmental conditions and complexities. To evaluate the efficacy of DEGE, we apply it to the domain of anomaly detection, rigorously testing the adapted DNNs within this specific context. Furthermore, we conduct comparative analyses between DEGE and established optimization methods using standard metrics to elucidate its advantages. Our findings shed light on DEGE’s effectiveness in addressing the challenges posed by dynamic environments, indicating its potential to revolutionize DNN optimization. As a practical application, we integrate DEGE-based DNNs into an IoT anomaly detection system to assess the overall impact of DEGE on anomaly detection performance. Our experiments demonstrate the efficiency of DEGE across 10 generations, showcasing its high adaptability to the dynamism inherent in IoT infrastructures. The proposed DEGE-based anomaly detection system processes highly dynamic environments within IoT infrastructure and classifies/predicts different types of anomalies efficiently with 99% detection accuracy across multiple benchmark and live experiment datasets. Solving the problem of multiclassification in dynamic abnormality detection, the proposed DEGE-based anomaly detection system was highly adaptable to the environment across generations, reaching the optimal DNN structure that delivers the best accuracy, and precision, with minimum loss value.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"167 ","pages":"Article 112377"},"PeriodicalIF":7.2,"publicationDate":"2024-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142703503","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep supervision network with contrastive learning for zero-shot sketch-based image retrieval","authors":"Zhenqiu Shu ,&nbsp;Guangyao Zhuo ,&nbsp;Jun Yu ,&nbsp;Zhengtao Yu","doi":"10.1016/j.asoc.2024.112474","DOIUrl":"10.1016/j.asoc.2024.112474","url":null,"abstract":"<div><div>Zero-shot sketch-based image retrieval (ZS-SBIR) is an extremely challenging cross-modal retrieval task. In ZS-SBIR, hand-drawn sketches are used as queries to retrieve corresponding natural images in zero-shot scenarios. Existing methods utilize diverse loss functions to guide deep neural networks (DNNs) to align feature representations of both sketches and images. In general, these methods supervise only the last layer of DNNs and then update each layer of DNNs using back-propagate technology. However, this strategy cannot effectively optimize the intermediate layers of DNNs, potentially hindering retrieval performance. To address this issue, we propose a deep supervision network with contrastive learning (DSNCL) approach for ZS-SBIR. Specifically, we employ a novel deep supervision network training method that attaches multiple projection heads to the intermediate layers of DNNs. These projection heads map multi-level features to a normalized embedding space and are trained by contrastive learning. The proposed method instructs the intermediate layers of DNNs to learn the invariance of various data augmentation, thereby aligning the feature representations of both sketches and images. This significantly narrows its domain gap and semantic gap. Besides, we use contrastive learning to directly optimize the intermediate layers of DNNs, which effectively reduces the optimization difficulty of their intermediate layers. Furthermore, we investigate the cross-batch metric (CBM) learning mechanism, which stores samples of different batches for metric learning by constructing a semantic queue, to further improve the performance in ZS-SBIR applications. Comprehensive experimental results on the Sketchy and TU-Berlin datasets validate the superiority of our DSNCL method over existing state-of-the-art methods.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"167 ","pages":"Article 112474"},"PeriodicalIF":7.2,"publicationDate":"2024-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142703497","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}