Graph convolutional neural networks (GCN) have shown the promise in medical image segmentation due to the flexibility of representing diverse range of image regions using graph nodes and propagating ...knowledge via graph edges. However, existing methods did not fully exploit the various attributes of image nodes and the context relationship among their attributes. We propose a new segmentation method with multi-similarity view enhancement and node attribute context learning (MNSeg). First, multiple views were formed by measuring the similarities among the image nodes, and MNSeg has a GCN based multi-view image node attribute learning (MAL) module to integrate various node attributes learnt from multiple similarity views. Each similarity view contains the specific similarities among all the image nodes, and it was integrated with the node attributes from all the channels to form the enhanced attributes of image nodes. Second, the context relationships among the attributes of image nodes are formulated by a transformer-based context relationship encoding (CRE) strategy to propagate these relationships across all the image nodes. During the transformer-based learning, the relationships were estimated based on the self-attention on all the image nodes, and then they were encoded into the learned node features. Finally, we design an attention at attribute category level (ACA) to discriminate and fuse the learnt diverse information from MAL, CRE, and the original node attributes. ACA identifies the more informative attribute categories by adaptively learn their importance. We validate the performance of MNSeg on a public lung tumor CT dataset and an in-house non-small cell lung cancer (NSCLC) dataset collected from the hospital. The segmentation results show that MNSeg outperformed the compared segmentation methods in terms of spatial overlap and the shape similarities. The ablation studies demonstrated the effectiveness of MAL, CRE, and ACA. The generalization ability of MNSeg was proved by the consistent improved segmentation performances using different 3D segmentation backbones.
•Image node attribute learning to integrate various node attributes from multiple similarity views.•The designed transformer-based strategy to encode context relationships of all image nodes.•The proposed attention to adaptively fuse the diverse node features and the original attributes.•The generalization ability of model demonstrated by utilizing several segmentation backbones.
ABSTRACT
A continuous attribute (e.g., calorie count) can be classified into separate categories (e.g., high vs. low), and a similar attribute value can fall into different categories depending on ...where the category boundaries are drawn. This research explores the effect of categorization on judgments of options (e.g., products and incentive‐compatible games) with continuous attributes. I predict and find a systematic preference shift between two options that were presented with different categorization criteria: When two options involve a tradeoff between two continuous attributes, people tend to prefer the option with both attributes classified into the favorable categories given the categorization criteria. I further show that this effect is driven by larger perceived differences between attribute values across category boundaries and is moderated by people's tendency to rely on category information. Overall, this effect holds even when people are highly familiar with the attributes and feel confident to make similarity evaluations, when people are cued that the categories provide little informational value, and when people are incentivized to make deliberate decisions. The findings in this research carry both theoretical and practical implications.
Rating norms for semantic attributes (e.g., concreteness, familiarity, valence) are widely used to study the content that people process as they encode meaningful material. Intensity ratings of ...individual attributes have been manipulated in numerous experiments with a range of memory paradigms, but those manipulations are contaminated by substantial correlations with the intensity ratings of other attributes. A method of controlling such contamination is needed, which requires a determination of how many distinct attributes there are among the large collection of attributes for which published norms are available. Identification of overlapping words in multiple rating projects yielded a data base containing normed values for each word’s perceived intensity (
M
rating) and ambiguity (rating
SD
) on 20 different attributes. Principal component analyses then revealed that the intensity space was spanned by just three latent semantic attributes, and the ambiguity space was spanned by five. Psychologically, the big three intensity factors (emotional valence, size, age) were highly interpretable, as were the big five ambiguity factors (discrete emotion, emotional valence, age, meaningfulness, and verbatim memory). We provide a data base of intensity and ambiguity factor scores that can be used to conduct uncontaminated studies of the memory effects of the intensity and ambiguity of latent semantic attributes.
Multi-authority attribute-based searchable encryption (MABSE) is an flexible and efficient way to securely share and search encrypted data. Compared with single-authority systems, MABSE has more ...complex access control policies and key management mechanism. However, most existing MABSE schemes rely on traditional number-theoretic assumptions, which maybe vulnerable to attack in the era of quantum-computers. Besides, the effective revocation of user attributes is also crucial in searchable encryption. To overcome these challenges, this paper proposes a new multi-authority ciphertext-policy attribute-based searchable encryption scheme for securely sharing encrypted data in the cloud. By calling Shamir’s threshold secret-sharing technology twice, we achieve co-management of the master key by attribute authorities and interactive generation of user private keys. Furthermore, the KUNodes algorithm is employed for attribute revocation, offering a mechanism to update private keys for non-revoked users. Compared to other schemes, MCP-ABSE-AR introduces multiple attribute authorities responsible for managing user attributes collectively. Additionally, it provides functionalities for keyword searching and attribute revocation. Finally, the proposed scheme is proved to be semantically secure under the decision learning with errors problem in the standard model.
•We annotate attribute labels on two large-scale person re-identification datasets.•We propose APR to improve re-ID by exploiting global and detailed information.•We introduce a module to leverage ...the correlation between attributes.•We speed-up the retrieval of re-ID by ten times with only a 2.92% accuracy drop.•We achieve competitive re-ID performance with the state-of-the-art methods.
Person re-identification (re-ID) and attribute recognition share a common target at learning pedestrian descriptions. Their difference consists in the granularity. Most existing re-ID methods only take identity labels of pedestrians into consideration. However, we find the attributes, containing detailed local descriptions, are beneficial in allowing the re-ID model to learn more discriminative feature representations. In this paper, based on the complementarity of attribute labels and ID labels, we propose an attribute-person recognition (APR) network, a multi-task network which learns a re-ID embedding and at the same time predicts pedestrian attributes. We manually annotate attribute labels for two large-scale re-ID datasets, and systematically investigate how person re-ID and attribute recognition benefit from each other. In addition, we re-weight the attribute predictions considering the dependencies and correlations among the attributes. The experimental results on two large-scale re-ID benchmarks demonstrate that by learning a more discriminative representation, APR achieves competitive re-ID performance compared with the state-of-the-art methods. We use APR to speed up the retrieval process by ten times with a minor accuracy drop of 2.92% on Market-1501. Besides, we also apply APR on the attribute recognition task and demonstrate improvement over the baselines.
•A δ-neighborhood relation on diversified attribute fuzzy decision information system.•A δ-neighborhood relation-based variable precision diversified multigranulation RFS.•A δ-neighborhood ...relation-based variable precision diversified multigranulation FRS.•A method of diversified attribute GDM problem with established model and VIKOR method.
For a considered multiple attribute group decision making (MAGDM) problem, there may be different evaluation attribute set used by different decision-makers for the same decision making problem. This article investigates the diversified attribute group decision making problem by using the diversified attribute multigranulation fuzzy rough set. Firstly, we give a general description and the definition of diversified attribute fuzzy decision making information system. We then construct the δ neighborhood according to the different attribute set over diversified attribute fuzzy decision making space. With the δ neighborhood relation classes, we establish the diversified attribute multigranulation decision making space and investigate the rough approximation of a crisp/fuzzy decision making object with respect to the diversified attribute multigranulation decision making space, i.e., the variable precision diversified attribute multigranulation fuzzy rough set model over diversified attribute fuzzy decision making space. Subsequently, we present some mainly conclusions and related mathematics properties for the established model. Furthermore, we explore the interrelationship between the proposed variable precision diversified attribute multigranulation fuzzy rough set model and the traditional models. Moreover, we propose a new approach to multiple attribute group decision making (MAGDM) problems with different evaluation attribute set based on variable precision diversified attribute multigranulation fuzzy rough set and VIKOR method. The background of the decision making problem, the principle and the methodology of the presented method are established as well as a particularly comparison of the existing approaches to multiple attribute group decision making with different evaluation attribute set shows the superiorities and deficiencies. At last, an numerical example of the performance appraisal for emergency preparedness of emergency management decision making is verified to illustrate the principle and procedures for this approach.
Attention has been widely paid to fuzzy β-covering, a powerful tool for processing uncertain information, and has recently turned into a hot topic. However, the ubiquity of unbalanced data makes ...fuzzy β-covering models are unavailable since they ignore the decision distribution normally. To address this drawback, a novel fuzzy β-covering attribute subset selection based on neighborhood composite entropy is proposed in this paper. First, the similarity measure of objects is discussed based on fuzzy β-neighborhoods for unbalanced fuzzy data. Subsequently, a neighborhood composite entropy is developed, which fully considers the decision distribution. Furthermore, an algorithm for attribute selection is presented according to the framework of the above measure mechanism. Furthermore, experimental results on 12 reality data from UCI and KRBM show that this novel method performs better classification on the proposed model to prove the performance and effectiveness by comparing several existing models.
Community search (CS) is an important research topic in network analysis, which aims to find a subgraph that satisfies the given conditions. A dynamic attribute heterogeneous information network ...(DAHIN) is a sequence of attribute heterogeneous information network (AHIN) snapshots, where each snapshot consists of multiple types of vertices as well as edges, and each vertex is associated a set of attribute keywords. CS over DAHIN faces many challenges. In this paper, we study the CS problem over DAHINs, aiming to search for cohesive subgraphs containing query vertex and simultaneously satisfying the connectivity, attribute cohesiveness and interaction stability. To this end, we propose a deep learning model with a three-level attention mechanism and the concept of interaction frequency with respect to multiple semantic relationships to measure the similarity of attributes and the stability of interactions between vertices respectively. In addition, we design three search algorithms to locate the target community by optimizing the degree of interaction stability and attribute similarity between vertices. Extensive experiments, including comparison with existing algorithms, ablation analysis, parameter sensitivity examination, and case studies, are conducted on four real-world datasets to validate the effectiveness and efficiency of the proposed model and search algorithms. The code and model of CS-DAHIN will be open source on GitHub.
•Identify blind spots of classification-based attribute reducts.•Propose a new notion of class-specific attribute reducts.•Provide a systematic study of class-specific attribute reducts.•Compare ...classification-based and class-specific attributes reducts.
The concept of attribute reducts plays a fundamental role in rough set analysis. There are at least two possibilities to define an attribute reduct. A classification-based or global attribute reduct is a minimal subset of condition attributes that preserves the positive region of the decision classification, namely, the positive regions of all decision classes, in a decision table. A class-specific, class-dependent, or local attribute reduct is a minimal subset of condition attributes that preserves the positive region of a particular decision class. While a classification-based reduct may not work equally well for every decision class, a class-specific attribute reduct is optimally tailored to a particular decision class. However, studies in rough set theory are dominated by classification-based reducts; there is very limited investigation on class-specific reducts. An objective of this paper is to draw attention to class-specific reducts. We systematically compare the two types of reducts and investigate their relationships with respect to both individual reducts and families of all reducts. It is possible to derive a class-specific reduct from a classification-based reduct and to derive a classification-based reduct from a family of class-specific reducts. The families of all class-specific reducts provide a pair of lower and upper bounds of the family of all classification-based reducts. Based on a three-way classification of attributes into the pair-wise disjoint sets of core, marginal, and nonuseful attributes, we examine relationships between the corresponding classes of classification-based and class-specific attributes. The union of the sets of class-specific core attributes is the set of classification-based core attributes. It is only possible to obtain an upper bound for the set of classification-based marginal attributes and a lower bound for the set of classification-based nonuseful attributes from the family of the class-specific corresponding sets of attributes.
In dynamic and open data environment, how to improve the performance of reduction is of great importance from incremental evaluation of attributes and quick search of attributes. In this paper, by ...considering both two perspectives, we first combine the incremental technology and the accelerated strategy in attribute reduction. On the one hand, we utilize the stable attribute group generated by DBSCAN to accelerate the process of searching reduction. On the other hand, we propose the matrix-based incremental mechanisms to dynamic attribute reduction when the objects are evolved over time. Moreover, these two methods are fused together in a unified algorithm of reduction. Finally, a series of comparative experiments is conducted to verify the effectiveness of proposed approach from stability, computational cost, and classification accuracy.
PDF
