Although fuzzy β-covering rough sets have emerged as a powerful extension of classical rough sets, certain existing models still lack a flexible multigranulation structure or fail to maintain the essential inclusion relation between lower and upper approximations. To address these deficiencies, this study introduces an advanced multigranulation fuzzy β-covering rough set model grounded in overlap and grouping functions. Crucially, the proposed model successfully guarantees the lower and upper inclusion relation. On this basis, a λ-multigranulation fuzzy β-neighborhood rough entropy measure is established as a comprehensive criterion integrating both algebraic and information-theoretic viewpoints. This measure is subsequently applied to construct a greedy attribute reduction algorithm. Extensive experiments on 15 benchmark datasets demonstrate that the proposed approach achieves significantly higher classification accuracy compared with five existing attribute reduction algorithms.

Achieving highly controlled and directional electron transfer at catalyst interface remains a persistent challenge in photocatalysis. Herein, Bi/Bi24O31Br10 nanosheets were prepared through a simple in situ reduction method to enhance the photocatalytic degradation of tetracycline (TC). Leveraging the surface plasmon resonance (SPR) effect of metallic Bi and the marked enhancement of interfacial charge transfer and separation efficiency, the optimized Bi/Bi24O31Br10 nanosheets exhibited significantly improved TC photodegradation performance, achieving a degradation efficiency of 80.78 % within 120 min under visible light irradiation-surpassing the performance of pristine Bi24O31Br10 nanosheets. Notably, the Bi/Bi24O31Br10 nanosheets demonstrated robust degradation efficiency under diverse conditions, including a broad pH range (3.0-9.0), varying TC concentrations (5-25 mg L-1), different photocatalyst dosages (5-20 mg), and in the presence of common inorganic anions (e.g., Cl-, HCO3-, SO42-, and NO3-). The mineralization process and degradation pathway of TC were elucidated through active sites trapping experiments, liquid chromatography-mass spectrometry (LC-MS), threedimensional excitation-emission matrix (3D EEM) spectroscopy, and density functional theory (DFT) calculations. Additionally, the ecotoxicity of TC and its degradation intermediates was evaluated using the Toxicity Estimation Software Tool (T.E.S.T.) and the Ecological Structure Activity Relationships (ECOSAR) model, revealing a reduction in environmental toxicity during the degradation process. This study provides valuable insights into the rational design of advanced Bi-based semiconductor photocatalysts for efficient environmental remediation.

Intuitionistic fuzzy information system (IFIS) is an extension of fuzzy information system that can represent more uncertain information and more accurately describe the essence of fuzziness. Attribute reduction is an important problem in processing and analyzing IFISs. The article tries to propose an attribute reduction method in view of intuitionistic fuzzy dominance mutual information in IFISs, whose information values are intuitionistic fuzzy numbers. First, an intuitionistic fuzzy dominance relation is established in IFISs according to intuitionistic fuzzy dominance degrees, and the intuitionistic fuzzy information structure generated by the intuitionistic fuzzy dominance relation is constructed. Then, the intuitionistic fuzzy dominance entropy and its variations are researched, and some of their properties are discussed. Subsequently, an attribute reduction method and its algorithm based on intuitionistic fuzzy dominance mutual information are given. Furthermore, numerical studies and statistical tests are presented to evaluate the performance of the proposed method. Theoretical research and experiments show that the raised attribute reduction method is applicable to IFISs.

Objective. The diagnosis of attention deficit hyperactivity disorder (ADHD) subtypes is important for the refined treatment of ADHD children. Although automated diagnosis methods based on machine learning are performed with structural and functional magnetic resonance imaging (sMRI and fMRI) data which have full observation of brains, they are not satisfactory with the accuracy of less than 80% for the ADHD subtype diagnosis. Approach. To improve the accuracy and obtain the biomarker of ADHD subtypes, we proposed a hierarchical binary hypothesis testing (H-BHT) framework by using brain functional connectivity (FC) as input bio-signals. The framework includes a two-stage procedure with a decision tree strategy and thus becomes suitable for the subtype classification. Also, typical FC is extracted in both two stages of identifying ADHD subtypes. That means the important FC is found out for the subtype recognition. Main results. We apply the proposed H-BHT framework to resting state fMRI datasets from ADHD-200 consortium. The results are achieved with the average accuracy 97.1% and an average kappa score 0.947. Discriminative FC between ADHD subtypes is found by comparing the P-values of typical FC. Significance. The proposed framework not only is an effective structure for ADHD subtype classification, but also provides useful reference for multiclass classification of mental disease subtypes.

Verbal and non-verbal human reaction generation is a challenging task, as different reactions could be appropriate for responding to the same behaviour. This paper proposes the first multiple and multimodal (verbal and nonverbal) appropriate human reaction generation framework that can generate appropriate and realistic human-style reactions (displayed in the form of synchronised text, audio and video streams) in response to an input user behaviour. This novel technique can be applied to various human-computer interaction scenarios by generating appropriate virtual agent/robot behaviours. Our demo is available at \url{https://github.com/SSYSteve/MRecGen}.

Abstract Background: Breast cancer (BC) is the second lethal cancer with the highest and rising morbidity in females. Emerging evidences have illustrated that circular RNAs (circRNAs) play essential roles in the tumorigenesis and metastasis of BC. However, the specific functions and underlying mechanistic involvement of circ_0000515 in BC have not yet been explored. Methods: Three BC datasets (GES101123, GES165884, and GES182471) from the NCBI GEO database were screened to identify differentially expressed circRNAs (DEcircRNAs). Then transwell and wound healing assays were performed to determinethe function of circ_0000515 in BC. The identification of downstream targets of circ_0000515 was performed using bioinformatics methods. RNA-pulldown assays, RIP assay, and CO-IP were further employed to identify the critical signaling pathway regulated by circ_0000515. Finally, rescue experiments were employed to confirm the connection between circ_0000515 and FUS in BC metastasis. Results: Circ_0000515 of a total of 49 DEcircRNAs was identified in BC datasets. Interestingly, the abundance of circ_0000515 was significantly increased in BC cells. Loss-of-functional experiments in vitro showed silencing circ_0000515 inhibited the invasion, migration and EMT process of BC. Mechanically, circ_0000515 stabilized the expression of FUS by impeding the interplay between FUS and MDM2, thereby protecting FUS from proteasomal degradation. Interestingly, we identified that FUS knockdown dramatically alleviated the promotive effect of circ_0000515 on BC metastasis. Conclusion: Circ_0000515 promoted invasion and migration of BC by attenuating MDM2-mediated FUS ubiquitination and degradation, and might function as a biomarker and therapeutic target for BC.

Some studies have shown that lyophilization significantly improves the stability of mRNA-LNPs and enables long-term storage at 2–8 °C. However, there is little research on the lyophilization process of mRNA-lipid nanoparticles (LNPs). Most previous studies have used empirical lyophilization with only a single lyoprotectant, resulting in low lyophilization efficiency, often requiring 40–100 h. In the present study, an efficient lyophilization method suitable for mRNA-LNPs was designed and optimized, shortening the total length of the lyophilization process to 8–18 h, which significantly reduced energy consumption and production costs. When the mixed lyoprotectant composed of sucrose, trehalose, and mannitol was added to mRNA-LNPs, the eutectic point and collapse temperature of the system were increased. The lyophilized product had a ginger root-shaped rigid structure with large porosity, which tolerated rapid temperature increases and efficiently removed water. In addition, the lyophilized mRNA-LNPs rapidly rehydrated and had good particle size distribution, encapsulation rate, and mRNA integrity. The lyophilized mRNA-LNPs were stable at 2–8 °C, and they did not reduce immunogenicity in vivo or in vitro. Molecular dynamics simulation was used to compare the phospholipid molecular layer with the lyoprotectant in aqueous and anhydrous environments to elucidate the mechanism of lyophilization to improve the stability of mRNA-LNPs. This efficient lyophilization platform significantly improves the accessibility of mRNA-LNPs.

We have carried out a long-timescale simulation study on crystal structures of nine antibody-antigen pairs, in antigen-bound and antibody-only forms, using molecular dynamics with enhanced sampling and an explicit water model to explore interface conformation and hydration. By combining atomic level simulation and replica exchange to enable full protein flexibility, we find significant numbers of bridging water molecules at the antibody-antigen interface. Additionally, a higher proportion of interactions excluding bulk waters and a lower degree of antigen bound CDR conformational sampling are correlated with higher antibody affinity. The CDR sampling supports enthalpically driven antibody binding, as opposed to entropically driven, in that the difference between antigen bound and unbound conformations do not correlate with affinity. We thus propose that interactions with waters and CDR sampling are aspects of the interface that may moderate antibody-antigen binding, and that explicit hydration and CDR flexibility should be considered to improve antibody affinity prediction and computational design workflows.

Attention Deficit Hyperactivity Disorder (ADHD) is a highly prevalent neurodevelopmental disease of school-age children. Early diagnosis is crucial for ADHD treatment, wherein its neurobiological diagnosis (or classification) is helpful and provides the objective evidence to clinicians. The existing ADHD classification methods suffer two problems, i.e., insufficient data and feature noise disturbance from other associated disorders. As an attempt to overcome these difficulties, a novel deep-learning classification architecture based on a binary hypothesis testing framework and a modified auto-encoding (AE) network is proposed in this paper. The binary hypothesis testing framework is introduced to cope with insufficient data of ADHD database. Brain functional connectivities (FCs) of test data (without seeing their labels) are incorporated during feature selection along with those of training data and affect the sequential deep learning procedure under binary hypotheses. On the other hand, the modified AE network is developed to capture more effective features from training data, such that the difference of inter- and intra-class variability scores between binary hypotheses can be enlarged and effectively alleviate the disturbance of feature noise. On the test of ADHD-200 database, our method significantly outperforms the existing classification methods. The average accuracy reaches 99.6% with the leave-one-out cross validation. Our method is also more robust and practically convenient for ADHD classification due to its uniform parameter setting across various datasets.

The expansion of three-dimensional protein structures and enhanced computing power have significantly facilitated our understanding of protein sequence/structure/function relationships. A challenge in structural genomics is to predict the function of uncharacterized proteins. Protein function deconvolution based on global sequence or structural homology is impracticable when a protein relates to no other proteins with known function, and in such cases, functional relationships can be established by detecting their local ligand binding site similarity. Here, we introduce a sequence order-independent comparison algorithm, PocketShape, for structural proteome-wide exploration of protein functional site by fully considering the geometry of the backbones, orientation of the sidechains, and physiochemical properties of the pocket-lining residues. PocketShape is efficient in distinguishing similar from dissimilar ligand binding site pairs by retrieving 99.3% of the similar pairs while rejecting 100% of the dissimilar pairs on a dataset containing 1538 binding site pairs. This method successfully classifies 83 enzyme structures with diverse functions into 12 clusters, which is highly in accordance with the actual structural classification of proteins classification. PocketShape also achieves superior performances than other methods in protein profiling based on experimental data. Potential new applications for representative SARS-CoV-2 drugs Remdesivir and 11a are predicted. The high accuracy and time-efficient characteristics of PocketShape will undoubtedly make it a promising complementary tool for proteome-wide protein function inference and drug repurposing study.

Model Agnostic Meta Learning (MAML) has become the most representative meta learning algorithm to solve few-shot learning problems. This paper mainly discusses MAML framework, focusing on the key problem of solving few-shot learning through meta learning. However, MAML is sensitive to the base model for the inner loop, and training instability occur during the training process, resulting in an increase of the training difficulty of the model in the process of training and verification process, causing degradation of model performance. In order to solve these problems, we propose a multi-stage loss optimization meta-learning algorithm. By discussing a learning mechanism for inner and outer loops, it improves the training stability and accelerates the convergence for the model. The generalization ability of MAML has been enhanced.

As generalizations of single-valued information systems, interval-valued information systems (IVISs) can better express the real data with uncertainty in some applications. Attribute reduction methods for complete IVISs or complete interval-valued decision systems (IVDSs) have been developed. However, there are few researches on attribute reduction for incomplete interval-valued information systems (IIVISs). The paper aims to investigate the attribute reduction issue in IIVISs. Firstly, the maximal and minimal distances, which characterize the difference between two interval values, are defined, and the maximal and minimal similarity degrees are given. Secondly, the fuzzy alpha-similarity relation is defined based on similarity between interval values, and the concept of alpha-equivalence relation is raised. Thirdly, entropy measures are investigated for IIVISs in view of alpha-equivalence relations. Fourthly, a new attribute reduction approach for IIVISs is proposed by using conditional entropy, and its corresponding algorithm is given. Finally, experiments to verify the effectiveness and feasibility of the newly proposed approach for attribute reduction in IIVISs are presented. These results will be helpful to perfect the uncertainty measurement model, and provide an approach for attribute reduction in IIVISs. (C) 2021 Published by Elsevier B.V.

Streaming processing of speech audio is required for many contemporary practical speech recognition tasks. Even with the large corpora of manually transcribed speech data available today, it is impossible for such corpora to cover adequately the long tail of linguistic content that's important for tasks such as open-ended dictation and voice search. We seek to address both the streaming and the tail recognition challenges by using a language model (LM) trained on unpaired text data to enhance the end-to-end (E2E) model. We extend shallow fusion and cold fusion approaches to streaming Recurrent Neural Network Transducer (RNNT), and also propose two new competitive fusion approaches that further enhance the RNNT architecture. Our results on multiple languages with varying training set sizes show that these fusion methods improve streaming RNNT performance through introducing extra linguistic features. Cold fusion works consistently better on streaming RNNT with up to a 8.5% WER improvement.

As generalizations of single-valued information systems, interval-valued information systems (IVISs) can better express real data. At present, numerous unsupervised attribute reduction approaches for single-valued information systems have been considered, but there are few researches on unsupervised attribute reduction for IVISs. In this article, we investigate a new fuzzy relation by means of similarity between interval values, and propose the concept of $$\alpha $$-approximate equal relation in view of the fuzzy similarity class. Then the equivalence relation induced by $$\alpha $$-approximate equal relation is used to define the information entropy, which is used to construct the unsupervised attribute reduction method together with mutual information for IVISs. Finally, experiments demonstrate that the advanced unsupervised attribute reduction method is effective and feasible in IVISs.

Kernel density estimation, which is a non-parametric method about estimating probability density distribution of random variables, has been used in feature selection. However, existing feature selection methods based on kernel density estimation seldom consider interval-valued data. Actually, interval-valued data exist widely. In this paper, a feature selection method based on kernel density estimation for interval-valued data is proposed. Firstly, the kernel function in kernel density estimation is defined for interval-valued data. Secondly, the interval-valued kernel density estimation probability structure is constructed by the defined kernel function, including kernel density estimation conditional probability, kernel density estimation joint probability and kernel density estimation posterior probability. Thirdly, kernel density estimation entropies for interval-valued data are proposed by the constructed probability structure, including information entropy, conditional entropy and joint entropy of kernel density estimation. Fourthly, we propose a feature selection approach based on kernel density estimation entropy. Moreover, we improve the proposed feature selection algorithm and propose a fast feature selection algorithm based on kernel density estimation entropy. Finally, comparative experiments are conducted from three perspectives of computing time, intuitive identifiability and classification performance to show the feasibility and the effectiveness of the proposed method.

In data processing, measurement of uncertainty is one of the significant evaluation tools, which can describe the uncertainty essence of data. So far, there are few measurable tools to study the uncertainty of type-2 fuzzy information systems (TFISs-2) (the expanded models of fuzzy information systems). This paper is devoted to looking for effective indicators to describe the uncertainty of TFISs-2. The fuzzy $$T_{\text {cos}}$$ -similarity relations are first introduced, which are generated by TFISs-2 based on Gaussian kernel. Then, the fuzzy information structures are defined on account of this fuzzy $$T_{\text {cos}}$$ -similarity relation. Next, two measures constructed from the upper and lower approximations are given for TFISs-2, that is, $$\delta$$ -accuracy and $$\delta$$ -roughness, which are used to reflect the degree of accuracy and inaccuracy of information by numerical form. Furthermore, by combining roughness and entropy, the $$\delta$$ -rough entropy is investigated. Finally, the practicability of proposed measures is tested by a numerical experiment. The experimental results show that the $$\delta$$ -rough entropy is efficacious and applicable for TFISs-2.

As one of the most common neurobehavioral diseases in school-age children, Attention Deficit Hyperactivity Disorder (ADHD) has been increasingly studied in recent years. But it is still a challenge problem to accurately identify ADHD patients from healthy persons. To address this issue, we propose a dual subspace classification algorithm by using individual resting-state Functional Connectivity (FC). In detail, two subspaces respectively containing ADHD and healthy control features, called as dual subspaces, are learned with several subspace measures, wherein a modified graph embedding measure is employed to enhance the intra-class relationship of these features. Therefore, given a subject (used as test data) with its FCs, the basic classification principle is to compare its projected component energy of FCs on each subspace and then predict the ADHD or control label according to the subspace with larger energy. However, this principle in practice works with low efficiency, since the dual subspaces are unstably obtained from ADHD databases of small size. Thereby, we present an ADHD classification framework by a binary hypothesis testing of test data. Here, the FCs of test data with its ADHD or control label hypothesis are employed in the discriminative FC selection of training data to promote the stability of dual subspaces. For each hypothesis, the dual subspaces are learned from the selected FCs of training data. The total projected energy of these FCs is also calculated on the subspaces. Sequentially, the energy comparison is carried out under the binary hypotheses. The ADHD or control label is finally predicted for test data with the hypothesis of larger total energy. In the experiments on ADHD-200 dataset, our method achieves a significant classification performance compared with several state-of-the-art machine learning and deep learning methods, where our accuracy is about 90 % for most of ADHD databases in the leave-one-out cross-validation test.

The uncertainty of information plays an important role in practical applications, so how to capture the uncertainty of information systems becomes more and more popular. Uncertainty measures can supply new viewpoints for processing information systems, and they can help us in disclosing the substantive characteristics of information. Fuzzy information systems are important research objects in artificial intelligence. As a special kind of fuzzy information system, fully fuzzy information system (FFIS) is worth studying. This article is devoted to search indicators for measuring uncertainty in a FFIS according to fuzzy information structures in view of Gaussian kernel, and the fuzzy information structures can be viewed as granular structures under granular computing. Firstly, by employing Gaussian kernel for calculating similarities among objects in a FFIS, the fuzzy Tcos-similarity relation is obtained. Then, based on this relation, fuzzy information structures in a FFIS are introduced. Next, according to the information structures, granulation measure of a given FFIS is advanced. Moreover, entropy measure is also considered for a given FFIS. Finally, two numerical experiments are conducted to interpret the realistic significance and potential applications for measuring uncertainty in a FFIS. Theoretical research, numerical experiments and validity analysis make clear that the proposed measures are efficacious and applicable for a FFIS.

Currently, there are no methods available offering solutions to select and identify antibodies binding to a specific conformational epitope of an antigen. Here, we developed a method to allow epitope-directed antibody selection from a phage display library by photocrosslinking bound antibodies to a site that specifically incorporates a noncanonical amino acid, p-benzoyl-l-phenylalanine (pBpa), on the target antigen epitope. By one or two rounds of panning against antibody phage display libraries, those hits that covalently bind to the proximity site of pBpa on specific epitopes of target antigens after ultraviolet irradiation are enriched and selected. This method was applied to specific epitopes on human interleukin-1β and complement 5a. In both cases, more than one-third of hits identified bind to the target epitopes, demonstrating the feasibility and versatility of this method.