The fuzziness index <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">m</i> has important influence on the clustering result of fuzzy algorithms, and it should not be forced to fix at usual value = 2. In view its distinctive features in applications limitation having 2 only, a recent advance called c-means with improved partitions (IFP-FCM) is extended this paper, generalized algorithm GIFP-FCM for more effective proposed. By introducing novel...

Generalizing out-of-distribution (OoD) is critical but challenging in real applications such as unmanned aerial vehicle (UAV) flight control. Previous machine learning-based control has shown promise dealing with complex real-world environments suffers huge performance degradation facing OoD scenarios, posing risks to the stability and safety of UAVs. In this paper, we found that introduced random noises during training surprisingly yield theoretically guaranteed performances via a proposed...

Ovarian cancer (OV) is regarded as one of the most lethal malignancies affecting female reproductive system, with individuals diagnosed OV often facing a dismal prognosis due to resistance chemotherapy and presence an immunosuppressive environment. T cells serve crucial mediator for immune surveillance elimination. This study aims analyze mechanism cell-associated markers in create prognostic model clinical use enhancing outcomes patients.

We introduce a method combining variational autoencoders (VAEs) and deep metric learning to perform Bayesian optimisation (BO) over high-dimensional structured input spaces. By adapting ideas from learning, we use label guidance the blackbox function structure VAE latent space, facilitating Gaussian process fit yielding improved BO performance. Importantly for problem settings, our operates in semi-supervised regimes where only few labelled data points are available. run experiments on three...

In this paper, a multiobjective evolutionary algorithm based soft subspace clustering, MOSSC, is proposed to simultaneously optimize the weighting within-cluster compactness and between-cluster separation incorporated within two different clustering validity criteria. The main advantage of MOSSC lies in fact that it effectively integrates merits good properties optimization-based approach for fuzzy clustering. This makes possible avoid trapping local minima thus obtain more stable results....

Aspect-level sentiment classification aims to predict the polarities towards target aspects given in sentences. To address issues of insufficient semantic information extraction and high computational complexity attention mechanisms existing aspect-level models based on deep learning, a contextual graph network (CGAT) is proposed. The proposed model adopts two networks aggregate syntactic structure into employs extract sentence-aspect sequences, aiming generate aspect-sensitive text...

Contrast enhancement plays an important role in image processing applications. This paper proposes a low-complexity automatic method for contrast enhancement. The exploits the high-frequency distribution of to estimate intensity-weighting matrix, which is then used control Gaussian fitting curve and shape gain. As such, can be easily designed enhance details hidden noteworthy regions. Subsequently, proposed grayscale transformation that obtained from rationally express distribution. Unlike...

Machine learning methods suffer from test-time performance degeneration when faced with out-of-distribution (OoD) data whose distribution is not necessarily the same as training distribution. Although a plethora of algorithms have been proposed to mitigate this issue, it has demonstrated that achieving better than ERM simultaneously on different types distributional shift datasets challenging for existing approaches. Besides, unknown how and what extent these work any OoD datum without...

With the rapid development of data applications in scene Industrial Internet Things (IIoT), how to schedule resources IIoT environment has become an urgent problem be solved.Due benefit its strong scalability and compatibility, Kubernetes been applied resource scheduling scenarios.However, limited types resources, default scoring strategy, lack delay control module limit performance.To address these problems, this paper proposes a multi-resource (MRS) scheme for IIoT.The MRS dynamically...

The COVID-19 pandemic brings the topic of citizen data management (CDAMA) into public eye. This study is one first attempts to analyze national approaches for CDAMA applied by governments different countries and continents in sectors. conducts a systematic overview representative contact tracing apps 21 four continents, collecting information aspects system. It then summarizes analyzes various governments’ system based on app overview. We found that priority between safety (i.e., health this...

Human Action Recognition (HAR) stands as a pivotal research domain in both computer vision and artificial intelligence, with RGB cameras dominating the preferred tool for investigation innovation this field. However, real-world applications, encounter numerous challenges, including light conditions, fast motion, privacy concerns. Consequently, bio-inspired event have garnered increasing attention due to their advantages of low energy consumption, high dynamic range, etc. Nevertheless, most...

摘要: 该文针对K平面聚类算法KPC (K-Plane Clustering)对噪声点敏感的缺陷，通过引入隶属度约束函数，推导出鲁棒的改进分割K平面聚类算法IFP-KPC(Improved Fuzzy Partitions for K-Plane Clustering)，并利用Voronoi距离对IFP-KPC算法的鲁棒性进行了合理解释。实验结果表明IFP-KPC算法较之于KPC算法具有更好的聚类效果。 关键词: K平面聚类; 改进模糊分割; Voronoi距离; 鲁棒性

In this paper, two novel soft subspace clustering algorithms, namely fuzzy weighting with competitive agglomeration (FWSCA) and entropy (EWSCA), are proposed to overcome the problems of unknown number clusters initialization prototypes for clustering. The main advantage FWSCA EWSCA lies in fact that they effectively integrate merits good properties agglomeration. This makes it possible obtain appropriate during progress. Moreover, algorithms can converge regardless initial initialization....

Sampled point and voxel methods are usually employed to downsample the dense events into sparse ones. After that, one popular way is leverage a graph model which treats points/voxels as nodes adopts neural networks (GNNs) learn representation of event data. Although good performance can be obtained, however, their results still limited mainly due two issues. (1) Existing GNNs generally adopt additional max (or mean) pooling layer summarize all node embeddings single graph-level for whole...

With an ever-greater increase in network bandwidth, speed and traffic, attack techniques are constantly changing improving, making it formidable for the traditional security defense measures to keep pace with this challenge. In paper, a theoretical analysis is made first of both intrusion detection data stream mining, then, research conducted into technique based on integration mining detection, thereby coming up algorithm light clustering through sliding damped window. And applied systems...

As a bio-inspired vision sensor, the spike camera emulates operational principles of fovea, compact retinal region, by employing discharges to encode accumulation per-pixel luminance intensity. Leveraging its high temporal resolution and neuromorphic design, holds significant promise for advancing computer applications. Saliency detection mimics behavior human beings captures most salient region from scenes. In this paper, we investigate visual saliency in continuous stream first time. To...

As a bio-inspired vision sensor, the spike camera emulates operational principles of fovea, compact retinal region, by employing discharges to encode accumulation per-pixel luminance intensity. Leveraging its high temporal resolution and neuromorphic design, holds significant promise for advancing computer applications. Saliency detection mimic behavior human beings capture most salient region from scenes. In this paper, we investigate visual saliency in continuous stream first time. To...

Power grid fault disposal preplan is an important reference for power disposal. Hence, extracting finegrained key entity information such as equipments, name and number from the basis computer to understand content further support intelligent A named recognition technology proposed based on deep learning. Firstly, character vector used represent text. Then features are extracted by combining attention mechanism bidirectional long short-term memory network. Finally, optimal serialization...

$k$ Nearest Neighbors ($k$NN) is one of the most widely used supervised learning algorithms to classify Gaussian distributed data, but it does not achieve good results when applied nonlinear manifold especially a very limited amount labeled samples are available. In this paper, we propose new graph-based $k$NN algorithm which can effectively handle both data and data. To goal, first constrained Tired Random Walk (TRW) by constructing an $R$-level nearest-neighbor strengthened tree over...

Purpose: Immediate X-ray examination is necessary while the surgical needle falls off during operation. In this study, one convolutional neural network (CNN) model was introduced for automatically detection in craniofacial images. Materials and Methods: The (5–0, ETHICON, USA) localized 8 different anatomic regions of 2 pig heads bilateral separately. Thirty-two images were obtained finally which cropped into fragmented divided training dataset test dataset. Then, immediate CNN developed...

By introducing a novel membership constraint function, new algorithm called fuzzy c-means switching regression model with generalized improved partitions (GIFP-FCRM) is proposed. This seems less sensitive to noise and outliers than the classical C (FCRM), provides fuzziness index m for (IFP-FCRM). Furthermore, parameter α, FCRM IFP-FCRM can be taken as two special cases of proposed algorithm. Several experimental results are presented demonstrate its advantage over in both insensitivity...