查看更多>>摘要：The electrical activities of the brain are recorded and measured with Electroencephalography (EEG) by means of placing the electrodes on the scalp of the brain. It is quite a famous and versatile methodology utilized in both clinical and academic research activities. In this work, sparse depiction is initially incorporated to the EEG signals by means of using K-Singular Value Decomposition (K-SVD) algorithm and the features are extracted by means of using Self-Organizing Map (SOM) technique. The extracted features are initially classified with Extreme Learning Machine (ELM) and the proposed classification versions of ELM such as Ensemble ELM model and Nature Inclined ELM Model. The proposed ensemble ELM model makes use of the combination of Modified Adaboost. RT based on wavelet thresholding with ELM. The proposed Nature Inclined ELM makes use of the combination of some famous swarm intelligence algorithms such as Genetic Algorithm based ELM (GA-ELM), Particle Swarm Optimization based ELM (PSO-ELM), Ant Colony Optimization based ELM (ACO-ELM), Artificial Bee Colony based ELM (ABC-ELM) and Glowworm Swarm Optimization based ELM (GSO-ELM). The extracted features are also classified with deep learning methodology by means of utilizing an incidental Long Short-Term Memory (LSTM) Recurrent Neural Network (RNN). Another famous methodology using Non-negative Matrix Factorization (NMF) and Affinity Propagation Congregation based Mutual Information (APCMI) with transfer learning techniques is also proposed and implemented once the sparse modelling is done and the results are analysed. The proposed methodology is implemented for two EEG datasets such as epilepsy dataset and schizophrenia dataset and a comprehensive analysis is done with very promising results.(C) 2022 Elsevier B.V. All rights reserved.
查看更多>>摘要：Histopathological image classification is of great importance in pathological diagnosis, such as tumor grading and cancer type identification. However, the traditional pathological examination is timeconsuming and requires the subjective judgments and rich experience of pathologists. In order to alleviate these problems and provide quantitative analysis results, this paper proposes a parallel 'same' and 'valid' convolutional block (PSV-CB) and an input-collaboration strategy for performing histopathological image classification. The core of PSV-CB is to employ different convolutional operations for learning hidden representations of each input respectively and then correspondingly integrate them together to highlight interesting contents, where an operational flow is constructed via multiple 'same' convolutions and followed by a max-pooling, which can be considered as a 'hard' feature coding. Another one is established using step-by-step 'valid' convolutions that consider feature extraction and down-sampling simultaneously, which can be regarded as a 'soft' feature coding. Therefore, the parallel 'same' and 'valid' convolutional neural network (PSV-ConvNet) is constructed using stacked PSV-CB according to the specific task. To compensate the loss of spatial information, we introduce an input-collaborative PSV-ConvNet (IC-PSV-ConvNet) that adds grayscale version of original inputs into the outputs of each PSV-CB for better fusing global information and learned features. The proposed IC-PSV-ConvNet is evaluated on three histopathological image datasets (lymphoma, breast cancer, and liver cancer) with satisfactory results. Our comparative experiments demonstrate that the proposed IC-PSV-ConvNet can achieve more accurate classification results compared to other existing ConvNets. (C)& nbsp;2022 Elsevier B.V. All rights reserved.
查看更多>>摘要：Butterfly protection is critical for environmental protection, and butterfly classification study is an essential tool for doing so. We proposed a new fine-grained butterfly classification architecture to address the issues of duplicate information in some butterfly images and trouble identifying them due to their tiny inter-class variance. To begin, a Non-Local Mean Filtering and Multi-Scale Retinex-based method (NL-MSR) is employed to enhance the butterfly images in order to efficiently retain more detail information. Then, to accomplish fine-grained categorization of butterfly images, a Multi-scale Sparse Network with Cross-Attention Mechanism (CA-MSNet) is designed. In CA-MSNet, a Cross-Attention Mechanism module (CAM) that offers distinct weights in the horizontal and vertical directions based on two strategies is devised to successfully identify the spatial distribution of butterfly stripes and spots and suppress incorrect information. Then, to overcome the recognition problem of butterfly spots with small inter-class variance, a Multi-scale sparse module (MSS) with multi-scale receptive fields is constructed. Finally, a Depthwise Separable Convolution module is employed to mitigate the parameter rise induced by the MSS network. In order to validate the model's feasibility and effectiveness in a complex environment, we compared it to existing methods, and our proposed method achieved an average recognition accuracy of 91.88%, with an F1 value of 92.15%, indicating that it has a good effect on the fine-grained classification of butterflies and can be applied to their recognition to realize their protection.(c) 2022 Elsevier B.V. All rights reserved.
查看更多>>摘要：The development of smart grids has enabled a wide variety of generation units to be included in the grid, which has made it necessary to predict the load demand to manage the grid accurately. In this study, a novel hybridization approach is proposed based on selecting the most appropriate method to be used in each prediction step. In this scope, Learning Vector Quantization (LVQ) is employed as a classifier while Elman Neural Networks (ENN), and Ridge Regression are selected as predictors. In the first stage, a forecasting model is built and trained with ENN and Ridge Regression models on training data using 1-h before, 2-h before, and the first-order derivative of electricity consumption data. In the second stage, the method by which the most successful results are obtained for each forecast is determined and labeled. Then the LVQ model is built and trained to determine the most accurate forecast by using the same input employed in modeling. Finally, forecasts are performed by deciding which model to be used with LVQ. Experimental results show that more accurate forecasting performance is obtained with the proposed approach than the other two individual models. The different combinations of conventional models are used to illustrate the effectiveness of the proposed selection strategy, and experimental results show that better forecasting performance is obtained with the proposed approach than the individual ones at each combination.(c) 2022 Elsevier B.V. All rights reserved.
查看更多>>摘要：One of the important issues facing HIS (Hospital Information System) in the context of big data is how to ensure that massive data and resources are protected from internal attacks and reduce the abuse of medical information. However, the existing single-value quantitative access control model based on trust or risk may not well reflect the true trust or risk situation because it cannot describe the timeliness and trend of the quantitative value. In response to this problem, we propose an access control model based on the credibility of the requesting user. Quantify the trust based on the user's historical visit records on the HIS, and introduce the user's historical behavior trend into the trust evaluation model through the corresponding regression analysis model. Comparative experiments show that in predicting linear, geometric, exponential, and mixed trends, the regression model in this paper is better than existing methods in predicting trust accuracy and predicting trust trends. Different from the working system of trusted access control model proposed in the existing literature, the model in this paper adds "Behavior warning module (BWM) ". The working mode that "User-visit, Early-warning, Trust-evaluation, Access-control "is very effective in reducing information leakage caused by visitors with non-profit purposes (such as curiosity) and purposeless (such as habitual browsing). And this also has a positive effect on improving the overall behavior level of users in the system. (C)& nbsp;2022 Elsevier B.V. All rights reserved.
查看更多>>摘要：In this paper, we propose a novel method called transcendental equation solver (TES) for solving transcendental equations. The TES comprises a generator defined by a neural network and a discriminator defined by the mathematical expression of the transcendental equation. First, a large amount of random noise is input into the TES generator to generate the solutions of the equation; subsequently, the solution is input into the discriminator and the discriminator calculates the error between the discriminator output and the true value. Moreover, this error can update the parameters in the generator with the backpropagation algorithm. The experimental results proved that the TES exhibits an improvement in accuracy, convergence speed, and stability compared to the other methods for solving transcendental equations. (C)& nbsp;2022 Elsevier B.V. All rights reserved.
查看更多>>摘要：With the increasing demands for low-delay network services, mobile edge computing (MEC) has emerged as an appealing solution to provide computing resources in close to the end users. Network function virtualization (NFV) is a new network architecture which replaces dedicated hardware middleboxes with software instances to run network functions via software virtualization on generalpurpose servers deployed at edge clouds. Because of the resource limitation at network edges, efficient placement and routing for online virtual network function requests (VNF-PRO) is a challenging task. The VNF-PRO has proven to be NP-hard, and thus, metaheuristic algorithms are the best choice in term of the solution quality. However, metaheuristics suffer from high computational complexity, and cannot be performed for online requests in the VNF-PRO. In this paper, a combined model based on fuzzy logic and genetic algorithm is proposed to achieve proper solution quality-speed trade-off in the VNF-PRO. In this method, a multi-criteria fuzzy inference system (named mcFIS) is used for the online VNF placement and routing. To achieve the best performance, a multi-objective evolutionary algorithm based on genetic algorithm (GA) is utilized in an offline procedure for automatic rule tuning of the mcFIS, once before applying it for online applications. Simulation results on two NFV benchmark instances demonstrate the efficiency of the proposed model against the existing techniques.(c) 2022 Elsevier B.V. All rights reserved.
查看更多>>摘要：Grey Wolf Optimizer (GWO) is one of the most important Swarm Intelligence based meta-heuristic algorithm which follows leadership mechanism and well planned hunting strategies among wolves' hierarchy. This paper has introduced a new variant of GWO termed as Randomized Balanced Grey Wolf Optimizer (RBGWO), which assists wolves to explore the search space in an efficient manner. The proposed algorithm improves the overall efficiency of the search process by establishing a balance between its exploitation and exploration capability incorporating three successive enhancement strategies equipped with social hierarchy mechanism and random walk with student's t-distributed random numbers. This newly proposed variant RBGWO has outperformed GWO and its other variants (RW-GWO, EGWO+ and EGWO*) in most of the cases on CEC 2014 benchmark functions with different scales. Results of the proposed variant have also been verified with the other meta-heuristic algorithms like GSA, CS, TPHS, CL-PSO, LX-BBO, B-BBO, SOS, DERand1Bin, Firefly, GWO, RW-GWO, EGWO+ and EGWO* on CEC 2014 benchmark functions. The statistical analysis of the results presents the efficiency of RBGWO (the proposed version) in overall performance. The state-of-the-art methods and the proposed algorithm have also been applied together to constrained and unconstrained real life problems. The results produced by the proposed variant are of better quality compared to that of others in these real-life problems also.(c) 2022 Elsevier B.V. All rights reserved.
查看更多>>摘要：The Capacitated Vehicle Routing Problem (CVRP) is a widely investigated NP-hard problem, which aims to determine the routes for a fleet of vehicles to serve a group of customers with minimum travel cost. In this paper, a fast evolutionary algorithm is proposed to solve CVRPs. To this end, a relevance matrix storing the probability that two customers are served successively by the same vehicle is calculated according to the local information of customer location and elite individuals in population. Based on the relevance matrix, an evolutionary algorithm called RMEA is proposed, where the relevance matrix is used to guide the crossover operation and accelerate the convergence of algorithm. Moreover, a relevance matrix based diversity preservation strategy is designed to increase the population diversity and solution quality. In the experiments, the proposed RMEA is compared to eight state-of-the-art heuristic methods tailored for CVRPs. Experimental results on three CVRP benchmarks demonstrate that the proposed RMEA is superior over eight compared algorithms and shows fast convergence speed. (C) 2022 Elsevier B.V. All rights reserved.
查看更多>>摘要：The assessment of the treatment technologies is a vital point for health care waste (HCW) man-agement. The assessment process is regarded as a multi-attribute group decision-making problem. Considering individual knowledge backgrounds and psychological preferences, and various uncertainty factors which are difficult to describe numerical and precisely, experts often portray their individual assessments by employing qualitative expressions. Linguistic term with weakened hedge (LTWH) is an effective tool to quantify the qualitative information. Additionally, the Bonferroni mean (BM) can capture the interrelationship between input data and the combined compromise solution (CoCoSo) method could produce reliable outcomes. In this paper, we extend the BM operator into the LTWH environment and propose the linguistic term with weakened hedge BM (LTWHBM) operator and the weighted version, i.e., LTWHWBM, to depict the indeterminacy in the HCW management and cope with assessment of the treatment technology problems based on the developed novel operational rules of LTWHs. Various properties and special cases of these operators are investigated in detail. Furthermore, we design a LTWHWBM based CoCoSo group decision-making model, herein the LTWHWBM operator is employed to integrate individual experts' preferences and the extend CoCoSo method is utilized to yield the ranking of alternatives. Ultimately, a case study concerning the assessment of the HCW treatment technologies is performed. The feasibility and favorable characteristics of the established model are justified by the comparisons with several prevailing methodologies and sensitivity analyses.(c) 2022 Elsevier B.V. All rights reserved.
NSTL
Elsevier