Information Sciences最新文献_第4页

Overlap Number of Balls Model-Agnostic CounterFactuals (ONB-MACF): A data-morphology-based counterfactual generation method for trustworthy artificial intelligence

IF 8.1 1区计算机科学

Information Sciences Pub Date : 2025-01-08 DOI: 10.1016/j.ins.2024.121844

José Daniel Pascual-Triana , Alberto Fernández , Javier Del Ser , Francisco Herrera

{"title":"Overlap Number of Balls Model-Agnostic CounterFactuals (ONB-MACF): A data-morphology-based counterfactual generation method for trustworthy artificial intelligence","authors":"José Daniel Pascual-Triana , Alberto Fernández , Javier Del Ser , Francisco Herrera","doi":"10.1016/j.ins.2024.121844","DOIUrl":"10.1016/j.ins.2024.121844","url":null,"abstract":"<div><div>Explainable Artificial Intelligence (XAI) is a pivotal research domain aimed at clarifying AI systems, particularly those considered “black boxes” due to their complex, opaque nature. XAI seeks to make these AI systems more understandable and trustworthy, providing insight into their decision-making processes. By producing clear and comprehensible explanations, XAI enables users, practitioners, and stakeholders to trust a model's decisions. This work analyses the value of data morphology strategies in generating counterfactual explanations. It introduces the Overlap Number of Balls Model-Agnostic CounterFactuals (ONB-MACF) method, a model-agnostic counterfactual generator that leverages data morphology to estimate a model's decision boundaries. The ONB-MACF method constructs hyperspheres in the data space whose covered points share a class, mapping the decision boundary. Counterfactuals are then generated by incrementally adjusting an instance's attributes towards the nearest alternate-class hypersphere, crossing the decision boundary with minimal modifications. By design, the ONB-MACF method generates feasible and sparse counterfactuals that follow the data distribution. Our comprehensive benchmark from a double perspective (quantitative and qualitative) shows that the ONB-MACF method outperforms existing state-of-the-art counterfactual generation methods across multiple quality metrics on diverse tabular datasets. This supports our hypothesis, showcasing the potential of data-morphology-based explainability strategies for trustworthy AI.</div></div>","PeriodicalId":51063,"journal":{"name":"Information Sciences","volume":"701 ","pages":"Article 121844"},"PeriodicalIF":8.1,"publicationDate":"2025-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143105393","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}

引用次数: 0

Inception MLP: A vision MLP backbone for multi-scale feature extraction

IF 8.1 1区计算机科学

Information Sciences Pub Date : 2025-01-08 DOI: 10.1016/j.ins.2024.121865

Jia Li , Rongchao Yang , Xinyan Cao , Bo Zeng , Zhandong Shi , Wei Ren , Xixin Cao

{"title":"Inception MLP: A vision MLP backbone for multi-scale feature extraction","authors":"Jia Li , Rongchao Yang , Xinyan Cao , Bo Zeng , Zhandong Shi , Wei Ren , Xixin Cao","doi":"10.1016/j.ins.2024.121865","DOIUrl":"10.1016/j.ins.2024.121865","url":null,"abstract":"<div><div>Recently, MLP-based networks have demonstrated remarkable performance in computer vision with simple but efficient structures. However, most existing MLP architectures struggle to balance the modeling of local and global regional information and often rely on static token mixing matrices for information fusion, disregarding the distinctiveness of different input contents. In this study, we propose inception MLP (iMLP), which employs multiple cross-MLP branches with varying receptive field sizes to simultaneously capture short-range and long-range dependencies. Meanwhile, the channel partition ratio <em>γ</em> is dynamically adjusted to better align with model characteristics as the network deepens. In addition, considering the diversity of input contents, we incorporate a lightweight, content-adaptive module to enable dynamic and efficient feature fusion. Experimental results demonstrate the versatility of iMLP as a competitive vision backbone across various visual tasks. For instance, our iMLP-S achieves 82.1% top-1 accuracy on the ImageNet-1K classification benchmark with only 20M parameters and extremely high throughput, outperforming state-of-the-art MLP-based models with a better trade-off between accuracy and computational efficiency.</div></div>","PeriodicalId":51063,"journal":{"name":"Information Sciences","volume":"701 ","pages":"Article 121865"},"PeriodicalIF":8.1,"publicationDate":"2025-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143105396","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}

引用次数: 0

Three-way concept lattice construction and association rule acquisition

IF 8.1 1区计算机科学

Information Sciences Pub Date : 2025-01-08 DOI: 10.1016/j.ins.2024.121867

Junping Xie , Jing Yang , Jinhai Li , Mingwei He , Huaxiang Song

引用次数: 0

Boosting occluded person re-identification by leveraging occlusion attributes

IF 8.1 1区计算机科学

Information Sciences Pub Date : 2025-01-08 DOI: 10.1016/j.ins.2024.121866

Tengfei Ren , Qiusheng Lian , Jiale Chen

{"title":"Boosting occluded person re-identification by leveraging occlusion attributes","authors":"Tengfei Ren , Qiusheng Lian , Jiale Chen","doi":"10.1016/j.ins.2024.121866","DOIUrl":"10.1016/j.ins.2024.121866","url":null,"abstract":"<div><div>Occluded person re-identification (ReID) aims to address the potential occlusion problem when matching occluded or holistic pedestrians from different camera views. Currently, occlusion augmentation-based methods have not fully exploited the occlusion attributes, resulting in suboptimal results. We introduce a novel ReID framework, dubbed Occlusion Attributes boosted Occluded Person Re-Identification (OA-ReID), aimed at leveraging the occlusion attributes for pedestrian-focused feature learning. Firstly, we propose an occlusion emulator (OE) that generates artificially occluded images towards emulating the occlusion scenarios. Both the original image and the corresponding artificially occluded image are jointly used for model training. Secondly, we present two crucial components, namely the inductive hard (IH) sample mining and the Occlusion-Informed Part Transformer (OIPT). The IH sample mining leverages the obstacle category to construct inductive triplets, which induces the model to extract identity-relevant features. The OIPT integrates the obstacle position information into our ReID framework to rectify the erroneous attention on occlusions, promoting reliable target pedestrian localization. Through extensive experiments, we show OA-ReID achieves state-of-the-art performance on both occluded and holistic person ReID benchmarks.</div></div>","PeriodicalId":51063,"journal":{"name":"Information Sciences","volume":"701 ","pages":"Article 121866"},"PeriodicalIF":8.1,"publicationDate":"2025-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143105410","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}

引用次数: 0

H∞ control for interval type-2 Takagi–Sugeno fuzzy systems via the membership-quadratic framework

IF 8.1 1区计算机科学

Information Sciences Pub Date : 2025-01-07 DOI: 10.1016/j.ins.2024.121839

KyungSoo Kim, PooGyeon Park

{"title":"H∞ control for interval type-2 Takagi–Sugeno fuzzy systems via the membership-quadratic framework","authors":"KyungSoo Kim, PooGyeon Park","doi":"10.1016/j.ins.2024.121839","DOIUrl":"10.1016/j.ins.2024.121839","url":null,"abstract":"<div><div>This paper aims to explore <span><math><msub><mrow><mi>H</mi></mrow><mrow><mo>∞</mo></mrow></msub></math></span> control synthesis for interval type-2 Takagi–Sugeno fuzzy systems using novel relaxation techniques. To tackle the conservatism of the previous works with respect to the utilization of membership properties, relaxation techniques are proposed within two distinct aspects. Based on a comprehensive interpretation of the properties of membership functions, an extrema-based algorithm to develop a convex polytope enclosing type-2 fuzzy membership distributions is proposed for membership-dependent stability analysis. To account for real-world conditions, this work utilizes constraints on the derivatives of lower and upper membership functions rather than the embedded membership function, in contrast to recent studies. In such a way as Finsler's lemma and relaxation principle, relaxation techniques that exploit lower, upper, and embedded membership properties with the reasonable constraints are established within a membership-quadratic framework with a matrix shrink lemma to save computational resources. As a result, the less conservative stabilization condition in the shape of the membership-quadratic framework is facilitated by the proposed algorithm and relaxation techniques. Finally, the effectiveness of the proposed methods is demonstrated through numerical examples.</div></div>","PeriodicalId":51063,"journal":{"name":"Information Sciences","volume":"701 ","pages":"Article 121839"},"PeriodicalIF":8.1,"publicationDate":"2025-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143105394","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}

引用次数: 0

Adaptive fault compensation for global performance tracking control of sensor faulty MIMO nonlinear systems with unmeasured states

IF 8.1 1区计算机科学

Information Sciences Pub Date : 2025-01-07 DOI: 10.1016/j.ins.2024.121862

Liuliu Zhang, Lingchen Zhu, Cheng Qian, Changchun Hua

{"title":"Adaptive fault compensation for global performance tracking control of sensor faulty MIMO nonlinear systems with unmeasured states","authors":"Liuliu Zhang, Lingchen Zhu, Cheng Qian, Changchun Hua","doi":"10.1016/j.ins.2024.121862","DOIUrl":"10.1016/j.ins.2024.121862","url":null,"abstract":"<div><div>In this article, the problem of global prescribed performance tracking control for multi-input multi-output (MIMO) nonlinear systems with sensor faults and unmeasured states is investigated. By constructing a state observer that incorporates an adaptive sensor fault compensation mechanism, the impact of the loss of sensor effectiveness is alleviated through the cubic absolute-value Lyapunov function analysis method. Based on several transformation functions and a time-varying scaling function, the tracking errors are restricted within the global prescribed performance without the constrained initial conditions. Considering the abrupt change in tracking errors due to the existence of sensor faults, a monitoring function to supervise the excessive loss of sensor effectiveness is designed. Furthermore, a novel reconfigurable controller can be constructed with the detected fault time instant and the global prescribed performance. The analysis demonstrates that all signals remain bounded and the tracking errors are maintained within the designed global prescribed performance regardless of sensor faults. Finally, the efficacy of the presented control scheme is demonstrated by the simulation results.</div></div>","PeriodicalId":51063,"journal":{"name":"Information Sciences","volume":"700 ","pages":"Article 121862"},"PeriodicalIF":8.1,"publicationDate":"2025-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143150774","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}

引用次数: 0

A semantics of the basic modal language based on a generalized rough set model

IF 8.1 1区计算机科学

Information Sciences Pub Date : 2025-01-07 DOI: 10.1016/j.ins.2024.121838

Md. Aquil Khan, Ranjan

引用次数: 0

A deep contrastive multi-modal encoder for multi-omics data integration and analysis

IF 8.1 1区计算机科学

Information Sciences Pub Date : 2025-01-07 DOI: 10.1016/j.ins.2024.121864

Ma Yinghua , Ahmad Khan , Yang Heng , Fiaz Gul Khan , Farman Ali , Yasser D. Al-Otaibi , Ali Kashif Bashir

{"title":"A deep contrastive multi-modal encoder for multi-omics data integration and analysis","authors":"Ma Yinghua , Ahmad Khan , Yang Heng , Fiaz Gul Khan , Farman Ali , Yasser D. Al-Otaibi , Ali Kashif Bashir","doi":"10.1016/j.ins.2024.121864","DOIUrl":"10.1016/j.ins.2024.121864","url":null,"abstract":"<div><div>Cancer is a highly complex and fatal disease that affects various human organs. Early and accurate cancer analysis is crucial for timely treatment, prognosis, and understanding of the disease's development. Recent research utilizes deep learning-based models to combine multi-omics data for tasks such as cancer classification, clustering, and survival prediction. However, these models often overlook interactions between different types of data, which leads to suboptimal performance. In this paper, we present a Contrastive Multi-Modal Encoder (CMME) that integrates and maps multi-omics data into a lower-dimensional latent space, enabling the model to better understand relationships between different data types. The challenging distribution and organization of the data into anchors, positive samples, and negative samples encourage the model to learn synergies among different modalities, pay attention to both strong and weak modalities, and avoid biased learning. The performance of the proposed model is evaluated on downstream tasks such as clustering, classification, and survival prediction. The CMME achieved an accuracy of 98.16% and an F1 score of 98.09% in classifying breast cancer subtypes. For clustering tasks across ten cancer types based on TCGA data, the adjusted Rand index reached 0.966. Additionally, survival analysis results highlighted significant differences in survival rates between different cancer subtypes. The comprehensive qualitative and quantitative results demonstrate that the proposed method outperforms existing methods.</div></div>","PeriodicalId":51063,"journal":{"name":"Information Sciences","volume":"700 ","pages":"Article 121864"},"PeriodicalIF":8.1,"publicationDate":"2025-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143150718","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}

引用次数: 0

Balancing convergence and diversity: Gaussian mixture models in adaptive weight vector strategies for multi-objective algorithms

IF 8.1 1区计算机科学

Information Sciences Pub Date : 2025-01-07 DOI: 10.1016/j.ins.2024.121858

Xuepeng Ren , Maocai Wang , Guangming Dai , Lei Peng , Xiaoyu Chen , Zhiming Song

{"title":"Balancing convergence and diversity: Gaussian mixture models in adaptive weight vector strategies for multi-objective algorithms","authors":"Xuepeng Ren , Maocai Wang , Guangming Dai , Lei Peng , Xiaoyu Chen , Zhiming Song","doi":"10.1016/j.ins.2024.121858","DOIUrl":"10.1016/j.ins.2024.121858","url":null,"abstract":"<div><div>In the study of decomposition-based multi-objective evolutionary algorithms, the adaptive weight vector approach effectively balances algorithm convergence and diversity. A common method for weight vector adaptation uses a population sparsity strategy, which calculates sparsity via Euclidean distance. However, this method causes individuals with low sparsity to cluster at the center of the objective space, while those with high sparsity spread to the edges, disrupting the convergence-diversity balance. To address this issue, this paper proposes using a Gaussian mixture model. This model treats data as a mix of multiple Gaussian distributions, partitioning the data space more flexibly. First, the paper analyzes various algorithms that adjust weight vectors using the sparsity strategy, highlighting their shortcomings. Then, it demonstrates how Gaussian mixture models can better divide the space and accurately identify individuals with different sparsity levels, correcting traditional sparsity calculation flaws. Since the population in the objective space changes during evolution, selecting appropriate component parameters is crucial. This paper uses the elbow rule to adaptively select these parameters. The experimental section includes three sets of experiments comparing the proposed algorithm with several popular algorithms, including a study on real mechanical bearing optimization. Results show that the proposed algorithm is highly competitive.</div></div>","PeriodicalId":51063,"journal":{"name":"Information Sciences","volume":"700 ","pages":"Article 121858"},"PeriodicalIF":8.1,"publicationDate":"2025-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143150770","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}

引用次数: 0

Differential evolution with individual and correlation information utilization for constrained optimization problems

IF 8.1 1区计算机科学

Information Sciences Pub Date : 2025-01-07 DOI: 10.1016/j.ins.2024.121863

Libao Deng , Guanyu Yuan , Chunlei Li , Lili Zhang

{"title":"Differential evolution with individual and correlation information utilization for constrained optimization problems","authors":"Libao Deng , Guanyu Yuan , Chunlei Li , Lili Zhang","doi":"10.1016/j.ins.2024.121863","DOIUrl":"10.1016/j.ins.2024.121863","url":null,"abstract":"<div><div>Compared to unconstrained global optimization, constrained optimization problems (COPs) introduce extra complexity to the search space due to constraints, making the issues more challenging to solve. COPs necessitate an algorithm with enhanced exploration and exploitation capabilities, as well as a constraint handling technique (CHT) that can be easily integrated and effectively balance constraints and objectives. To fully leverage evolutionary information in solving COPs, this paper proposes a constrained differential evolution algorithm (called IUCDE) based on the utilization of individual and correlation information. The constraints, objectives, and distances of individuals in the population are integrated into individual information, which is represented by individual performance scores. Based on the individual information, individuals are categorized into three subpopulations with distinct search characteristics to maximize their potential. Based on the correlation information generated in each iteration of the population, a dynamic feasibility rule is proposed, which, combined with the original feasibility rule, is adaptively selected to handle constraints based on the proportion of feasible solutions in the population. The proposed IUCDE algorithm is compared with five state-of-the-art constrained optimization algorithms across 22 test problems from the CEC 2017 benchmark, demonstrating superior performance. Furthermore, IUCDE exhibits a competitive advantage in solving 41 test problems from the CEC 2020 real-world constrained optimization test benchmark. Extensive experiments have validated the efficient execution of IUCDE and the effectiveness of its components.</div></div>","PeriodicalId":51063,"journal":{"name":"Information Sciences","volume":"700 ","pages":"Article 121863"},"PeriodicalIF":8.1,"publicationDate":"2025-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143150772","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}

引用次数: 0