查看更多>>摘要：Nowadays, more and more optimization techniques are used to deal with complex engineering optimization problems. Firefly algorithm (FA) inspired by the flash communication between fireflies, has been proven to be competitive with other swarm intelligence algorithms and has been widely applied to solve complex engineering optimization problems. However, FA has some defects in dealing with complex engineering optimization problems, such as the exploration and exploitation cannot be well balanced. Therefore, in order to achieve effective performance, the different characteristics of search strategies can be applied at different stages of the search process to achieve a balance between exploration and exploitation. In this paper, a multi-strategy firefly algorithm with selective ensemble (MSEFA) is proposed. In MSEFA, the algorithm has three novel search strategies with different characteristics in the strategy pool. In addition, an idea of selective ensemble is adopted to design a priority roulette selection method. The method can select suitable search strategies in different search stages and coordinate the balance of strategies so that better results can be obtained. Furthermore, a parameter adaptive transformation mechanism is designed to control the decreasing rate of step size alpha. To verify the effectiveness of MSEFA, performance tests are conducted on the CEC 2013 and CEC 2019 test suites, after which MSEFA is used to solve four complex engineering optimization problems. Experimental results show that MSEFA has the best performance compared with other FA variants and other improved swarm intelligence algorithms. In addition, MSEFA also achieves the best results in dealing with four complex engineering optimization problems. (c) 2022 Elsevier B.V. All rights reserved.
查看更多>>摘要：Accurate prediction of blood glucose (BG) is conducive to avoiding abnormal blood glucose events and improving blood glucose management for Type 1 diabetes (T1D) patients. Recently, many deep learning-based methods for BG prediction have been proposed with encouraging results. However, most deep prediction methods do not consider the time-dependent scale discrepancy of different variables on BG dynamics and use the same time window for all input variables. This neglect will directly lead to information redundancy on short-term related variables or information incompletion on long-term related variables, which is not conducive to prediction accuracy. In this regard, we proposed an autonomous channel deep learning framework for personalized multivariate BG prediction. The autonomous channel network in the proposed framework learns representation from input variables with reasonable sampling periods and sequence lengths based on the domain knowledge of time-dependent scale between variables, thereby effectively avoiding input information redundancy and incompletion. The framework was evaluated on a clinical dataset, OhioT1DM Dataset, with experimental results in terms of root mean square error (RMSE) (18.930 +/- 2.155 mg/dL) with the mean absolute relative difference (MARD) (9.218 +/- 1.607%) for prediction horizons (PH) = 30 min. These are the best-reported results for BG prediction when compared with other methods including the support vector regression (SVR), the long short-term memory network (LSTM), the dilated recurrent neural network (DRNN), the temporal convolutional networks (TCN), and the deep residual time-series forecasting (DRTF). (c) 2022 Elsevier B.V. All rights reserved.
查看更多>>摘要：Crime linkage is a difficult task and is of great significance to maintaining social security. It can be treated as a binary classification problem. Some crimes are difficult to determine whether they are serial crimes under the existing evidence, so the two-way decisions are easy to make mistakes for some case pairs. Here, the three-way decisions based on the decision-theoretic rough set are applied and its key issue is to determine thresholds by setting appropriate loss functions. However, sometimes the loss functions are difficult to obtain. In this paper, a method to automatically learn thresholds of the three-way decisions without the need to preset explicit loss functions is proposed. We simplify the loss function matrix according to the characteristic of crime linkage, re-express thresholds by loss functions, and investigate the relationship between overall decision cost and the size of the boundary region. The trade-off between the uncertainty of the boundary region and the decision cost is taken as the optimization objective. We apply multiple traditional classification algorithms as base classifiers, and employ real-world cases and some public datasets to evaluate the effect of our proposed method. The results show that the proposed method can reduce classification errors. (c) 2022 Elsevier B.V. All rights reserved.
查看更多>>摘要：Differential evolution (DE) algorithm has attracted considerable attention because of its effectiveness and simplicity. However, previous studies have validated that DE still suffers from some limitations such as premature convergence and slow convergence especially dealing with multimodal optimization problems. To address these concerning issues, we propose an innovative optimization method named particle swarm-differential evolution algorithm with multiple random mutation (MRPSODE) in this paper. The proposed MRPSODE algorithm is based on multiple random mutation framework cooperating with mean particle swarm mutation strategy, DE/current-to-rand/1 mutation strategy and disturbance strategy. Firstly, we incorporate the modified mean particle swarm mutation strategy into DE algorithm to improve the global convergence ability. Secondly, DE/current-to-rand/1 mutation strategy is adopted to increase the population diversity and produce perturbations to avoid the algorithm trapping into a local optimum. Thirdly, we propose a disturbance strategy to help the population escape from local optima, so as to enhance the exploration ability. Finally, to ensure that the proposed algorithm can get satisfactory solutions with a fast convergence speed, we design a multiple random mutation framework, in which these three mutation strategies can effectively play their advantages and make up for the shortcomings of others. To evaluate the performance of the proposed algorithm, three different experiments are constructed on twenty-nine classical benchmark functions. The simulation results demonstrate that, (1) MRPSODE significantly outperforms conventional PSO and DE algorithms, (2) MRPSODE can achieve better performance than nine well-known DE variants in terms of solution quality and robustness, (3) MRPSODE is superior to nine latest heuristic-based algorithms. Furthermore, MRPSODE is successfully applied to seven typical constrained optimization problems and performs better than almost all compared methods. (C) 2022 Elsevier B.V. All rights reserved.
查看更多>>摘要：In wireless sensor and actuator networks (WSANs), sensing region partition, task allocation, and collector schedule greatly affect the performance of WSANs. Specifically, nonuniform data aggregation causes unbalanced energy consumption, thereby reducing the network lifetime. Additionally, unbalanced allocation of data collection tasks of collectors increases the gathering delay. Thus, it is critical to balance the energy consumption of sensor nodes, as well as task allocation and scheduling with the limited number of collectors. This study proposes a novel splitting-merging-based automatic scheduling scheme to balance energy consumption and task allocation for WSANs. First, a sensing region splitting-merging scheme is proposed to make the data load uniform, in which the sensing region is self-organized into many subregions taking into account the maximum number of sensor nodes and the maximum distance between sensor nodes in each subregion. Second, a task allocation strategy based on genetic algorithm is proposed to optimize the number of collectors and uniformly assign tasks to collectors. Finally, the trajectory of each collector is optimized by applying ant colony optimization algorithm. Numerical experiments show that the proposed method outperforms well in terms of prolonging the network lifetime and reducing the gathering delay when compared with several related methods. (C)& nbsp;2022 The Author(s). Published by Elsevier B.V.
查看更多>>摘要：The increasing free access of the Internet provides us with favorable circumstances to investigate search engine index reflecting more and more personal behavior information. Part of valuable travel search information can assist us to achieve more robust and reliable prediction of metro passenger flow. Inspired by this, the paper develops a new multi-source time series fusion and direct interval prediction approach to grasp the dynamic law of metro passenger flow effectively. Multi-source index regarding metro travel from three major search engines (Baidu, Sogou and 360) in China are screened out and fused into the powerful predictors. By integrating an optimized multivariate mode decomposition strategy and long short-term memory model, lower and upper bounds of prediction interval are estimated directly by a multi-objective framework that combines the advantages of both the deep learning models long short-term memory and the ensemble learning approach. Especially, two sets of real experiment data of Beijing and Shanghai metro systems are employed to test our approach. Findings show that fusion of multi-source index information promotes the predictability of metro passenger flow, contributing to improving operation management and service quality. (C) 2022 Elsevier B.V. All rights reserved.
查看更多>>摘要：Active suspension systems in road vehicles are applied in order to mitigate the road-induced chassis vertical accelerations more effectively than standard passive suspensions, thus increasing comfort and handling. Such systems are greatly assisted by road preview schemes, consisting of special sensors usually based on laser scanners (e.g. LiDAR sensors), which detect road irregularities ahead of the vehicle and feed this information to a control system, designed to manipulate the active suspension accordingly. In this paper, a model predictive controller (MPC) with road preview incorporating radial basis function (RBF) models, is presented as a control scheme for a full car active suspension system. The employed RBF models can efficiently approximate the nonlinear behavior of the suspension system, thus improving performance over linear MPC methods. Special care is taken to alleviate the increased computational complexity entailed in the RBF models, in order to ensure that online implementation of the controller is feasible. The proposed scheme is evaluated on a detailed simulated full car model under various road excitation types, while making use of a realistic approach for incorporating LiDAR road scanner noise. Comparisons to a passive suspension system, as well as a standard MPC controller with a fully linear plant model, demonstrate the performance potential of using RBF prediction models in a road preview MPC context. (c) 2022 Elsevier B.V. All rights reserved.
查看更多>>摘要：Ant Colony Optimization (ACO) is a family of nature-inspired metaheuristics often applied to finding approximate solutions to difficult optimization problems. Despite being significantly faster than exact methods, the ACOs can still be prohibitively slow, especially if compared to basic problem-specific heuristics. As recent research has shown, it is possible to significantly improve the performance through algorithm refinements and careful parallel implementation benefiting from multi-core CPUs and dedicated accelerators. In this paper, we present a novel ACO variant, namely the Focused ACO (FACO). One of the core elements of the FACO is a mechanism for controlling the number of differences between a newly constructed and a selected previous solution. The mechanism results in a more focused search process, allowing to find improvements while preserving the quality of the existing solution. An additional benefit is a more efficient integration with a problem-specific local search. Computational study based on a range of the Traveling Salesman Problem instances shows that the FACO outperforms the state-of-the-art ACOs when solving large TSP instances. Specifically, the FACO required less than an hour of an 8-core commodity CPU time to find high-quality solutions (within 1% from the best-known results) for TSP Art Instances ranging from 100 000 to 200 000 nodes. (c) 2022 Elsevier B.V. All rights reserved.
查看更多>>摘要：Autism spectrum disorder (ASD) is a lifelong neurological condition that affects how a person interacts and learns. The early and accurate diagnosis of ASD is vital to developing a comprehensive rehabilita-tion plan that improves the quality of life and the integration of the ASD person in the social, family, and work environment. However, the accurate diagnosis of ASD is usually affected since it is linked to the judgment of an expert, which produces biases related to the lack of objectivity. In consequence, several works have been dedicated to developing early detection techniques for ASD based on Machine Learning (ML) technologies and eye-tracking tools. The present work aims to introduce a new methodology for ASD classification, which uses Kernel Extreme Learning Machine (KELM), an objective dataset based on gaze tracking, feature extraction techniques, and data augmentation for training the model. In turn, to enhance the accuracy in ASD classification, the KELM model is optimized through the Giza Pyramids Construction (GPC) algorithm. The proposed approach includes pipeline data augmentation, dimensionality reduction, and a posterior normalization to classify ASD subjects accurately. Statistical tests and analyses were performed to validate the performance of the proposed methodology, resulting in an average accuracy of 98.8% in ASD classification. (c) 2022 Elsevier B.V. All rights reserved.
Patwary, Muhammed J. A.Cao, WeipengWang, Xi-ZhaoHaque, Mohammad Ahsanul...
11页
查看更多>>摘要：Automatic recognition of bedridden patients' physical activity has important applications in the clinical process. Such recognition tasks are usually accomplished on visual data captured by RGB, depth, and/or thermal cameras by utilizing supervised machine learning. However, supervised machine learning requires a large amount of labeled data and the accuracy depends on extracting appropriate features based on the domain knowledge. A plausible solution to these issues is using semi-supervised learning that focuses less on labeled data and domain knowledge. In this paper, a novel fuzziness-based semi supervised multimodal learning algorithm, called (FSSL-PAR) is proposed for bedridden patient activity recognition. We use a synergistic interaction on RGB, Depth, and Thermal videos to assess the effect of visual multimodality for the first time in this semi-supervised learning setting. Experiments are conducted on a dataset collected by mimicking the patients with Acute Brain Injury (ABI) from a neurorehabilitation center. The results exhibit the superiority of the proposed algorithm over the existing supervised learning algorithms. We also see a positive correlation between the performance and the size of the labeled data in the proposed FSSL-PAR. (c) 2022 Elsevier B.V. All rights reserved.
NSTL
Elsevier