Expert Systems with Applications最新文献

{"title":"A high-effective swarm intelligence-based multi-robot cooperation method for target searching in unknown hazardous environments","authors":"","doi":"10.1016/j.eswa.2024.125609","DOIUrl":"10.1016/j.eswa.2024.125609","url":null,"abstract":"<div><div>To solve target searching problems for multi-robot cooperation with inaccurate target distance perception in unknown hazardous environments, a hybrid adaptive robotic particle swarm optimizer (RPSO) and grey wolf optimizer (GWO) based algorithm with continuous virtual target guidance is proposed for high effective path planning in the searching. In the initial searching stages, both the wolf behavior-generated position and the <em>gbest</em> position and the <em>pbesti</em> positions from RPSO are employed to guide the motions of robots. With the information provided by these initial robot movement paths, a geometric model is established to generate potential targets. The K-means cluster algorithm is introduced to estimate a virtual target position online from potential targets, with new robot-presenting route information to update the history path information. Then the virtual position is employed as one of the direction components to help the robots approach the actual target. In addition, to avoid mobile robots falling into local convergence, a heuristic moving direction determination scheme is utilized to make robots circumvent obstacles in swarm motions, as well as a mutual repulsion algorithm to keep them in a scattering state. Simulation experiments on different types of unknown environments with varied robot numbers and adaptability testing for a dynamic target are carried out to verify the feasibility of the proposed target searching method with comparisons to the other three famous target searching algorithms. It is verified from the results that the presented method can not only contribute a 100% success rate in all runs of searching for a stochastic dynamic target under a limited maximal velocity, but also produce both the shortest path length and minimum iterations in terms of statistical metrics over the comparative methods.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":null,"pages":null},"PeriodicalIF":7.5,"publicationDate":"2024-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142571809","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":"On the local convergence of ADAM-DPGAN with simultaneous and alternating gradient decent training methods","authors":"","doi":"10.1016/j.eswa.2024.125646","DOIUrl":"10.1016/j.eswa.2024.125646","url":null,"abstract":"<div><div>Generative Adversarial Networks (GANs) do not ensure the privacy of the training datasets and may memorize sensitive details. To maintain privacy of data during inference, various privacy-preserving GAN mechanisms have been proposed. Despite the different approaches and their characteristics, advantages, and disadvantages, there is a lack of a systematic review on them. This paper first presents a comprehensive survey on privacy-preserving mechanisms and offers a taxonomy based on their characteristics. The survey reveals that many of these mechanisms modify the GAN learning algorithm to enhance privacy, highlighting the need for theoretical and empirical analysis of the impact of these modifications on GAN convergence. Among the surveyed methods, ADAM-DPGAN is a promising approach that ensures differential privacy in GANs for both the discriminator and the generator networks when using the ADAM optimizer, by introducing appropriate noise based on the global sensitivity of discriminator parameters. Therefore, this paper conducts a theoretical and empirical analysis of the convergence of ADAM-DPGAN. In the presented theoretical analysis, assuming that simultaneous/alternating gradient descent method with ADAM optimizer converges locally to a fixed point and its operator is L-Lipschitz with L &lt; 1, the effect of ADAM-DPGAN-based noise disturbance on local convergence is investigated and an upper bound for the convergence rate is provided. The analysis highlights the significant impact of differential privacy parameters, the number of training iterations, the discriminator’s learning rate, and the ADAM hyper-parameters on the convergence rate. The theoretical analysis is further validated through empirical analysis. Both theoretical and empirical analyses reveal that a stronger privacy guarantee leads to a slower convergence, highlighting the trade-off between privacy and performance. The findings also indicate that there exists an optimal value for the number of training iterations regarding the privacy needs. The optimal settings for each parameter are calculated and outlined in the paper.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":null,"pages":null},"PeriodicalIF":7.5,"publicationDate":"2024-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142593552","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":"How to assess measurement capabilities of a security monitoring infrastructure and plan investment through a graph-based approach","authors":"","doi":"10.1016/j.eswa.2024.125623","DOIUrl":"10.1016/j.eswa.2024.125623","url":null,"abstract":"<div><div>Security monitoring is a crucial activity in managing cybersecurity for any organization, as it plays a foundational role in various security processes and systems, such as risk identification and threat detection. To be effective, security monitoring is currently implemented by orchestrating multiple data sources to provide corrective actions promptly. Poor monitoring management can compromise an organization’s cybersecurity posture and waste resources. This issue is further exacerbated by the fact that monitoring infrastructures are typically managed with a limited resource budget. This paper addresses the problem of supporting security experts in managing security infrastructures efficiently and effectively by considering the trade-off cost-benefit between using specific monitoring tools and the benefit of including them in the organization’s infrastructure. To this aim, we introduce a graph-based model named <em>Metric Graph Model</em> (MGM) to represent dependencies between security metrics and the monitoring infrastructure. It is used to solve a set of security monitoring problems: (i) <em>Metrics Computability</em>, to assess the measurement capabilities of the monitoring infrastructure, (ii) <em>Instrument Redundancy</em>, to assess the utility of the instruments used for the monitoring, and (iii) <em>Cost-Bounded Constraint</em>, to identify the optimal monitoring infrastructure in terms of cost-benefit trade-off. We prove the NP-hardness of some of these problems, propose heuristics for solving them based on the Metric Graph Model and provide an experimental evaluation that shows their better performance than existing solutions. Finally, we present a usage scenario based on an instance of the Metric Graph Model derived from a state-of-the-art security metric taxonomy currently employed by organizations. It demonstrates how the proposed approach supports an administrator in optimizing the security monitoring infrastructure in terms of saving resources and speeding up the decision-making process.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":null,"pages":null},"PeriodicalIF":7.5,"publicationDate":"2024-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142586042","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep reinforcement learning-based multi-objective optimization for electricity–gas–heat integrated energy systems","authors":"","doi":"10.1016/j.eswa.2024.125558","DOIUrl":"10.1016/j.eswa.2024.125558","url":null,"abstract":"<div><div>With the increasing global attention on energy efficiency and carbon emissions, the optimization of integrated energy systems (IES) has become the key to improve energy efficiency and reduce pollution emissions. However, most of the existing optimization methods cannot effectively deal with the complexity of high dimensional continuous action space. Therefore, this paper focuses on a novel multi-objective optimization strategy for the electricity–gas–heat integrated energy systems (EGH-IES). Firstly, considering the absorption capacity of wind power and the emission of pollutant gases, a multi-objective optimization model is constructed based on the mechanism model and operation constraints of each device in EGH-IES, in which the integrated operation cost and the environmental factors are taken as optimization objectives. Then, the multi-objective optimization problem is designed as the optimal strategy of interaction learning between agent and environment in reinforcement learning, and the output power of the devices constitutes the action of reinforcement learning. Additionally, the Ornstein–Uhlenbeck process is introduced to enhance the training efficiency and exploration performance of the agent, and the deep deterministic policy gradients (DDPG) algorithm is employed to optimize the action, thus the output power of the appliances could be obtained. Finally, the simulation results show that compared with deep Q network (DQN) method and proximal policy optimization (PPO) method, the reward function value of the proposed method increases by 2.43% and 6.09%, respectively, which represents a reduction in economic cost and pollutant emissions. These verify the effectiveness and superiority of the proposed multi-objective optimization scheme in cost reduction and benefit improvement for the EGH-IES.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":null,"pages":null},"PeriodicalIF":7.5,"publicationDate":"2024-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142586289","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":"Acoustic fault diagnosis of three-phase induction motors using smartphone and deep learning","authors":"","doi":"10.1016/j.eswa.2024.125633","DOIUrl":"10.1016/j.eswa.2024.125633","url":null,"abstract":"<div><div>Faults in induction motors can halt production lines in factories, leading to downtime and resulting in production and economic losses. Therefore, it is crucial to ensure that motors operate reliably. This paper describes an approach for the acoustic fault diagnosis of rotor bars in three-phase induction motors (IM). The authors analyzed the following conditions: a healthy IM, an IM with one broken rotor bar, an IM with two broken rotor bars, and an IM with three broken rotor bars. The FFT method was used to compute the FFT spectrum of the acoustic signals. An original feature extraction method DWV (Differences of Word Vectors) was proposed to compute the acoustic features. DenseNet-201, ResNet-18, ResNet-50, and EfficientNet-b0 were used to classify these acoustic features. The computed recognition efficiency is 100 %. The proposed method was also verified using a low-pass filter of 1–1225 Hz and word coding.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":null,"pages":null},"PeriodicalIF":7.5,"publicationDate":"2024-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142593548","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":"Design of adaptive recommendation system for autism children using optimal feature selection-based adaptive dilated 1DCNN-LSTM with attention mechanism","authors":"","doi":"10.1016/j.eswa.2024.125399","DOIUrl":"10.1016/j.eswa.2024.125399","url":null,"abstract":"<div><div>One kind of neurological disorder is caused in the brain which is defined as Autism Spectrum Disorder (ASD). It has acquired the symptoms that appear in young children. In addition to that, it influences how the individual behaves and learns as well as communicates and interacts with others. More specifically, the term Autism is defined as a developmental disorder that impacts communication and social skills and it may vary from mental handicap cases to relieving superior cognitive abilities, intact, and the characteristic pattern of poor. Moreover, the school activities have acquired various difficulties to the given model that include changes in expected routines, intense sensory stimulation, noisy or disordered environments, and social interactions. Consequently, the conventional approaches face certain limitations like user privacy, scalability, and cold-start. Here, a novel suggestion system for autistic children is developed to detect distractions and anxious situations using deep learning and then treat the children based on their abilities. It has helped to prevent the risk to children. The data is given to the selection of the feature stage. The weight optimization is performed using the Modified Garter Snake Optimization Algorithm (MGSOA) during the selection of features. Then, the selected features are given to the Adaptive Dilated One Dimensional Conventional Neural Network (1DCNN) and Long Short-Term Memory (LSTM) with Attention Mechanism termed AD-1DCNN + LSTM-AMfor detecting the autism disorder for children. Here, the parameter optimization is performed using MGSOA optimization. It effectively forecasts the symptoms in a short time. This optimization helps to provide reliable and flexible outcomes for the developed recommendation system for autistic children. The developed recommendation system for autistic children is compared to baseline techniques with efficacy metrics to visualize elevated results.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":null,"pages":null},"PeriodicalIF":7.5,"publicationDate":"2024-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142552635","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":"LMSFF: Lightweight multi-scale feature fusion network for image recognition under resource-constrained environments","authors":"","doi":"10.1016/j.eswa.2024.125584","DOIUrl":"10.1016/j.eswa.2024.125584","url":null,"abstract":"<div><div>In many resource-constrained environments, recognition tasks often require efficient and fast execution. Currently, many methods designed for this field adopt a combination of convolutional operations and Vision Transformers (ViTs) to achieve comprehensive feature representation while maintaining efficient performance. However, these methods still have higher parameter counts or floating point operations (FLOPs), making it difficult to adapt more resource-constrained environments. Therefore, a lightweight Multi-Scale Feature Fusion Network (LMSFF) is proposed to address this issue. The proposed method mainly consists of three modules: lightweight local processing (LLP) modules, local–global fusion modules (LGFM), and lightweight information fusion (LIF) modules. The LLP modules, considering the issue of computational redundancy, propose a branch structure that effectively reduces parameter consumption while maintaining high performance. To capture more comprehensive contextual information, the LGFM fuses local and global features, thus enhancing the comprehensive representation of image features. The LIF extracts crucial features through pooling operations at different scales while preserving lightweight characteristics. Additionally, to enhance the model’s generalization, a new weighted loss function is introduced, which alleviates the long-tail distribution issue in real-world scenarios and improves recognition performance for rare categories. Experimental results demonstrate that LMSFF achieves better balance between recognition accuracy and resource consumption compared with other state-of-the-art lightweight hybrid models.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":null,"pages":null},"PeriodicalIF":7.5,"publicationDate":"2024-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142539026","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":"Competitive multi-task Bayesian optimization with an application in hyperparameter tuning of additive manufacturing","authors":"","doi":"10.1016/j.eswa.2024.125618","DOIUrl":"10.1016/j.eswa.2024.125618","url":null,"abstract":"<div><div>Multi-task Bayesian optimization is an effective approach for optimization involving multiple correlated tasks. Typically, either all the tasks or one primary task should be optimized, depending on the objectives of the problems. We consider optimizing the primary task without explicitly pre-determining which is the primary task. Instead, the primary task is defined as the task whose optimal value is the best among all tasks. Due to the black-box nature of the tasks, the decision makers are not able to identify the primary task beforehand. It is thus critical for the algorithms to recognize and optimize the true primary task. Such problems are called competitive multi-task problems and arise in areas including machine learning and engineering design. In this work, we propose a competitive multi-task Bayesian optimization (CMTBO) algorithm to solve competitive multi-task problems. It selects the query point as well as the task to query in each optimization iteration. We theoretically analyze the regret bounds for the algorithm and test their performances on several synthetic and real-world problems. In addition, our algorithm is applied to a material extrusion (an important technology in additive manufacturing) problem to tune the process parameters and select material types.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":null,"pages":null},"PeriodicalIF":7.5,"publicationDate":"2024-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142586043","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 efficient Q-learning integrated multi-objective hyper-heuristic approach for hybrid flow shop scheduling problems with lot streaming","authors":"","doi":"10.1016/j.eswa.2024.125616","DOIUrl":"10.1016/j.eswa.2024.125616","url":null,"abstract":"<div><div>Efficient scheduling in flow shop environments with lot streaming remains a critical challenge in various industrial settings, necessitating innovative approaches to optimize production processes. This study investigates a hybrid flow shop scheduling problem dominant in real-world printed circuit board assembly shops. A novel multi-objective hyper-heuristic combining Q-learning, i.e., two-stage improved spider monkey optimization (TS-ISMO), is tailored to address the complexities of the flow shop scheduling problems. The proposed method aims to simultaneously optimize conflicting objectives such as minimizing makespan, total energy consumption, and total tardiness time while incorporating lot streaming considerations. For multi-objective hyper-heuristic techniques, the algorithm dynamically selects and adapts a diverse set of low-level heuristics to explore the solution space comprehensively and strike a balance among competing objectives. The proposed TS-ISMO algorithm incorporates several significant features aimed at enhancing its performance. These features encompass hybrid heuristics for solution initialization, a contribution value method for comprehensive convergence and diversity assessment, diverse evolutionary state judgments to promote the algorithm’s balance between exploration and exploitation capabilities, and a Q-learning strategy for self-adaptive parameter tuning. The integration of Q-learning facilitates intelligent parameter control, enabling the algorithm to autonomously adjust its behavior based on past experiences and evolution dynamics. This adaptive mechanism enhances convergence speed and solution quality by effectively guiding the search process toward promising regions of the solution space. Extensive computational experiments are conducted on benchmark instances of hybrid flow shop scheduling problems with lot streaming to evaluate the performance of the proposed algorithm. Comparative analyses against state-of-the-art approaches demonstrate its superior solution quality and computational efficiency.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":null,"pages":null},"PeriodicalIF":7.5,"publicationDate":"2024-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142538955","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":"Explainable artificial intelligence-based framework for efficient content placement in elastic optical networks","authors":"","doi":"10.1016/j.eswa.2024.125541","DOIUrl":"10.1016/j.eswa.2024.125541","url":null,"abstract":"<div><div>The rapid development of telecommunication networks brings new optimization problems and the urgent need for dedicated and highly efficient solution methods. Recently, the idea of aiding network optimization with machine learning (<span>ml</span>) algorithms has gained more and more attention in the research society. Despite numerous successful applications of these methods, their adaption in real networks and systems is hindered due to the lack of a full explainability of their decisions and, in turn — the lack of trust. Hopefully, these aspects may be addressed by explainable artificial intelligence methods (<span>xai</span>). In this paper, we study an essential problem of the anycast content placement. Having a set of physical data centers (<span>dc</span>s) located in selected network nodes and a set of different contents (services), the task is to decide in which <span>dc</span>s place each of the contents in order to improve the optical network performance (measured as a bandwidth blocking probability (<span>bbp</span>)). To this end, we propose a dedicated <span>ml</span>-based framework, which approaches the placement problem as a supervised learning task of predicting network’s <span>bbp</span> for a content placement configuration. We perform extensive numerical experiments to tune the framework, considering five supervised learning algorithms and three comparison metrics. We also use explainable artificial intelligence methods to interpret the models’ decisions and draw general conclusions regarding beneficial content placement in a real network. Lastly, we compare the performance of the proposed <span>ml</span>-based placement framework with three reference methods. The results prove our approach’s extremely high efficiency, which reduced the <span>bbp</span> significantly compared to the best reference approach. Depending on the network settings and the offered traffic volume, the framework allowed to serve up to 47% of the traffic that would be rejected by the best reference method (corresponding to 3.76 Tbps of data).</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":null,"pages":null},"PeriodicalIF":7.5,"publicationDate":"2024-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142538956","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}