查看更多>>摘要：In this paper, we propose the q-rung orthopair fuzzy weighted averaging (q-ROFWA) aggre-gation operator (AO) of q-rung orthopair fuzzy numbers (q-ROFNs). The proposed q- ROFWA AO of q-ROFNs can overcome the drawbacks of the q-rung orthopair fuzzy interac-tion weighted Hamy mean (q-ROFIWHM) AO and the q-rung orthopair fuzzy power weighted Maclaurian symmetric mean (q-ROFPWMSM) AO of q-ROFNs. Moreover, we pro -pose a new group decision making (GDM) method based on the proposed q-ROFWA AO of q-ROFNs. The proposed GDM method can overcome the drawbacks of the existing GDM methods. It provides us a very useful approach to deal with GDM problems in q-rung ortho-pair fuzzy environments. (c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：Recently, visual trackers based on region proposal networks (RPN) have attracted widespread attention due to their relatively high efficiency and excellent performance. RPNbased trackers mainly combine a classification branch and a regression branch to predict a target's state. These branches are all under the guidance of pre-defined anchor boxes. RPN-based trackers, however, first compute the Intersection-over-Union (IoU) between the anchor boxes and ground truth boxes, and then use a fixed IoU threshold to separate negative and positive training samples. The limit of this design lies in the fact that these trackers lack an analysis of the actual content of the intersecting regions, which may include distractor objects or few meaningful regions of the tracked target. In this research, we propose a probabilistic anchor assignment with region proposal network (PaaRPN) that can adaptively separate anchors into negative samples and positive samples according to the model's current learning status. To this end, we first calculate the classification scores of the anchor boxes conditioned on the current model and fit a probability distribution to the classification scores. The whole tracking model is then trained with anchor boxes separated into negative and positive samples in a probabilistic manner. Moreover, we introduce an online learning method in the PaaRPN framework that enables the model to have powerful discriminative abilities by exploiting both background and target appearance information. We tested the PaaRPN tracker on six tracking benchmarks to exhibit the effectiveness of the proposed method. In particular, our model outperforms a strong RPN tracker, SiamRPN++, with AUC scores improvements of 0.613 ! 0.657 and 0.496 ! 0.565 on UAV123 and LaSOT, respectively. (c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：Both regularized multi-task learning (RMTL) and direct multi-task twin support vector machine (DMTSVM) have shown good performances in dealing with multi-task problems. They all use hyperplanes to realize classification. However, the hyperplane cannot reflect the distribution of data well. Therefore, in this paper, we propose a novel multi-task twin hypersphere support vector machine (MTTHSVM) to solve multi-task classification prob-lems. It will generate two hyperspheres rather than hyperplanes for each task. So, the pro-posed method could better describe the distribution information of all training samples compared with the existing RMTL and DMTSVM. Based on Hierarchical Bayes theory, MTTHSVM divides the center of each hypersphere into task-specific and task-common parts to better measure the commonality and individuality of tasks. Then the shared infor-mation contained in multiple related tasks could be adaptively mined well. Therefore, the prediction accuracy will be improved to some extent. Besides, MTTHSVM is superior to RMTL and DMTSVM in terms of computational efficiency. This is because our MTTHSVM just solves two smaller-sized quadratic programming problems without any matrix inverse operations. Experimental results on one artificial data set, thirty-five benchmark data sets and a real image data set Cifar100 have verified the effectiveness of our method.(c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：Network intrusion detection is one of the most important components of mobile networks security. In recent years, the application of neural networks has been very popular in network intrusion detection. However, due to limited resources of IoT devices, fast detection of the intrusion requires a high accuracy neural network with a lightweight and efficient architecture. Therefore, the conventional architectures of neural networks are not suitable for intrusion detection in IoT devices due to the use of a large number of parameters in these models concerning the limited processing resources in IoT devices. This paper presents a new and lightweight architecture based on Parallel Deep Auto-Encoder (PDAE) that uses both locally and surrounding information around individual values in the feature vector. This type of separation of features allows us to increase the accuracy of the model while greatly reducing the number of parameters, memory footprint, and the need for processing power. The effectiveness of the proposed model is evaluated using KDDCup99, CICIDS2017, and UNSW-NB15 datasets and the results shows the superiority of the proposed model over the state-of-the-art algorithms in terms of both accuracy and performance.(c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：In practice, stock market behavior is difficult to predict accurately because of its high volatility. To improve market forecasts, a method inspired by Elman neural network and quantum mechanics is presented. To render the network sensitive to dynamic information, the internal self-connection signal that is extremely useful for system modeling is introduced to the proposed technique. Double chains quantum genetic algorithm is employed to tune the learning rates. This model is validated by forecasting closing prices of six stock markets, the simulation results indicate that the proposed algorithm is feasible and effective. Accordingly, generalizing the method is deemed advantageous.(c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：Genesis and geometrical properties of fuzzy line are explored in this study. The existing studies on fuzzy geometry suggest that fuzzy lines can be characterized as a union of crisp lines with varying membership values but this characterization is not capable to identify the nature of the fuzzy point situated on the fuzzy line. Expressing fuzzy line as a collection of fuzzy points in the geometrical plane is an intended goal of this paper. In the first part, a fuzzy line has been interpreted as a fuzzification function that produces a collection of fuzzy numbers. The equivalency between the union of lines with different membership grades and collection of fuzzy numbers is also shown. In the next part, the possible nature of a fuzzy points situated on a generic fuzzy line is explored. Analogous to the concept of a line in classical geometry, this paper constructs and approximates a fuzzy line as a collec-tion of fuzzy points. All the proposals and results are illustrated numerically and geomet-rically in a fuzzy geometrical plane.(c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：Despite the rapid development of support vector regression (SVR), it costs unacceptable training time in large-scale datasets and is hard to fit complex, high frequency oscillating, and non-stationary time series data. SVRs are still perplexed by the selection of critical parameters and hidden noise in input data. This work proposes a hybrid model to overcome these issues that need to be resolved, namely EEMD-GRU-TWSVRCSSA. The proposed model utilizes twin support vector regression (TWSVR) to overcome the shortcomings of the SVR in terms of training time and fitting accuracy. A novel meta-heuristic algorithm, cloud salp swarm algorithm (CSSA), is employed to automatically select the optimal hyper parameters for the TWSVR. The ensemble empirical mode decomposition (EEMD) reduces the influences of hidden noise in the input data, meanwhile splitting the high-frequency and low-frequency sub-datasets and feeding them to the gated recurrent unit (GRU) and TWSVR-based model, respectively. The forecasting of the proposed algorithm and other alternative algorithms are conducted on three real-world electric load datasets from the National Electricity Market (NEM), Queensland and New South Wales regions, Australia, and the well-known National Grid UK. Experimental results demonstrate the superiority and competitiveness of the proposed algorithm. (c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：Optimization design of product structures is a critical point of intelligent manufacturing that is often overlooked, particularly their performance balance involving different uncertainties. It is urgent to connect product consumers, design experts, and manufacturing plants to balance the optimization of product structures. As an input of intelligent manufacturing, an industrial network based on software-defined networking (SND) for product optimization design including structure performance balance is established in this study, which breaks down the barriers between an industrial network and the performance balance optimization of product structures. Accordingly, the high-dimensional coupling clustering of product components is performed through a two-dimensional (2D) plane mapping with a design structure matrix (DSM) that aims at the entire product life cycle (PLC). A random chance-constrained programming model for the performance balance optimization of product structures is developed, where the fuzzy expert evaluation is considered to include heterogeneous uncertainties, and an integrated multi-objective discrete cuckoo algorithm nested with Monte Carlo simulation is designed to solve the model. The rationality and superiority of the proposed method are verified using a case study with a large hydraulic machine tool for sheet stamping where product modules with better comprehensive performance are generated.(c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：In this paper, the fully distributed event-triggered time-varying formation control of heterogeneous linear multiagent systems is studied. To make the reference trajectory controllable and flexible, the bounded input is introduced to leader agent. The intention of this paper is to decrease the unnecessary information transmission of agents via communication topology by designing the intermittent transmission control protocol. Firstly, the adaptive state compensator is designed to evaluate the state information of leader agent based on the estimation value of leader state and two kinds of event-triggered mechanisms. Then, two types of output-feedback time-varying formation controllers are presented based on different formation feasible conditions. By theoretical analysis, the estimation error of event-triggered state compensator is proved to be uniformly ultimately bounded and the event-triggered mechanisms prevent Zeno behavior, and the formation tracking errors converge to the small region of zero. Finally, some simulation experiments are provided to support the theoretical results.(c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：Video-based facial expression recognition (FER) has received increased attention as a result of its widespread applications. However, a video often contains many redundant and irrel-evant frames. How to reduce redundancy and complexity of the available information and extract the most relevant information to facial expression in video sequences is a challeng-ing task. In this paper, we divide a video into several short clips for processing and propose a clip-aware emotion-rich feature learning network (CEFLNet) for robust video-based FER. Our proposed CEFLNet identifies the emotional intensity expressed in each short clip in a video and obtains clip-aware emotion-rich representations. Specifically, CEFLNet con-structs a clip-based feature encoder (CFE) with two-cascaded self-attention and local-glo-bal relation learning, aiming to encode clip-based spatio-temporal features from the clips of a video. An emotional intensity activation network (EIAN) is devised to generate emo-tional activation maps for locating the salient emotion clips and obtaining clip-aware emotion-rich representations, which are used for expression classification. The effective-ness and robustness of the proposed CEFLNet are evaluated using four public facial expres-sion video datasets, including BU-3DFE, MMI, AFEW, and DFEW. Extensive experiments demonstrate the improved performance of our proposed CEFLNet in comparison with the state-of-the-art methods. (c) 2022 Elsevier Inc. All rights reserved.
NSTL
Elsevier