查看更多>>摘要：Infectious diseases usually have the characteristics of rapid spread with a large impact range. Once they break out, they will cause a large area of infection, which creates tremendous health and security risks. Thus, early infectious disease monitoring and prevention are critical. Current surveillance systems can predict the incidence of infectious diseases to a certain extent. However, the diversity, inaccuracy and incompleteness of the data collected by sensors make it difficult to obtain accurate monitoring results. Moreover, the limited local resources of a monitoring system cannot process the increasing volume of data in a timely manner. To address these challenges, fuzzy logic and edge computing have been applied to infectious disease monitoring in recent years. This paper presents a comprehensive review of infectious disease monitoring technologies based on fuzzy logic and edge computing. Fuzzy neural networks in infectious disease surveillance are introduced in detail, followed by a brief study of applications of fuzzy systems in infectious disease surveillance. Finally, improvements in existing disease detection systems based on the combination of edge computing and fuzzy logic are described. The review shows that edge computing and fuzzy logic are complementary and that their combination greatly improves the processing efficiency and the storage space of the data. At the same time, with edge computing as the carrier, the combination of fuzzy logic, neural networks, expert systems and other technologies can effectively carry out disease prediction and diagnosis.
查看更多>>摘要：Imbalanced learning is important and challenging since the problem of the classification of imbalanced datasets is prevalent in machine learning and data mining fields. Sampling approaches are proposed to address this issue, and cluster-based oversampling methods have shown great potential as they aim to simultaneously tackle between-class and within-class imbalance issues. However, all existing clustering methods are based on a one-time approach. Due to the lack of a priori knowledge, improper setting of the number of clusters often exists, which leads to poor clustering performance. Besides, the existing methods are likely to generate noisy instances. To solve these problems, this paper proposes a deep instance envelope network-based imbalanced learning algorithm with the multilayer fuzzy c-means (MlFCM) and a minimum interlayer discrepancy mechanism based on the maximum mean discrepancy (MIDMD). This algorithm can guarantee high quality balanced instances using a deep instance envelope network in the absence of prior knowledge. First, the MlFCM is designed for the original minority class instances to obtain deep instances and increase the diversity of instances. Then, the MIDMD is proposed to avoid the generation of noisy instances and maintain the consistency of the interlayers of instances. Next, the multilayer FCM and minimum interlayer discrepancy mechanism are combined to construct a deep instance envelope network – the MlFC&IDMD. Finally, an imbalance learning algorithm is proposed based on the MlFC&IDMD. In the experimental section, thirty-three popular public datasets are used for verification, and over ten representative algorithms are used for comparison. The experimental results show that the proposed approach significantly outperforms other popular methods.
查看更多>>摘要：Generalized Fuzzy Hyperbolic Models (GFHMs) offer a simpler structure and less computational complexity than the typical fuzzy systems. Type-2 fuzzy systems, in contrast, have better handling of uncertainty but at the cost of higher computational complexity. Here, we propose a synergistic hybrid framework of interval type-2 fuzzy systems and GFHMs for a better uncertainty handling and simpler computational structure in the modelling and control of nonlinear systems. For this purpose, we first extend the GFHM to a computational model with various width and subsequently propose interval type-2 generalized fuzzy hyperbolic systems (IT2-GFHS) as a computational framework for nonlinear systems modelling. We then employ this IT2-GFHS in a general sliding-based robust nonlinear controller. Theoretical Lyapunov analysis reveals the overall asymptotic stability of the resulting closed-loop system. The numerical simulations for system modelling and identification on two nonlinear benchmark problems also reveal higher accuracy, lower computation time, and fewer adjustable parameters for the proposed IT2-GFHS models. Furthermore, applications to two nonlinear benchmark control problems show similar performance in terms of robustness to noise and disturbances compared with type-2 fuzzy systems, with the IT2-GFHS-based nonlinear controller having considerably fewer computations and floating-point operations. Finally, the proposed approach is experimentally implemented to control a 3-Prismatic-Series-Prismatic (3-PSP) parallel robot. Experimental results also confirm the improved tracking performance of the proposed method compared with interval type-2 and type-1 fuzzy systems, while also requiring fewer adjustable parameters.
查看更多>>摘要：Aiming at the attitude control problem of hypersonic reentry vehicles (HRVs), a deep reinforcement learning (DRL) based anti-disturbance control method is proposed. First, a compound control framework consisting of a DRL-based auxiliary controller and a fixed-time anti-disturbance controller is proposed to improve the control performance under the premise of ensuring stability. Then, a novel value function approximation mechanism, named experience-based value expansion (EVE), is proposed to modify the value function update equation based on a two-dimensional replay buffer, which solves the DRL convergence problem brought by the HRV's strong nonlinearities, tight coupling, and big flight envelope. Furthermore, a result-oriented encoder (ROE) is proposed to solve the DRL generalization problem brought by the HRV's high uncertainties and unavailable real training environment. A bottleneck shape neural network structure is used for the DRL's network structure to extract high-dimensional features and prevent overfitting to the training environment. Finally, abundant numerical comparative simulations demonstrate the effectiveness of the proposed efficient DRL algorithms and the DRL-based attitude controller.
查看更多>>摘要：The growing popularity of quantitative trading in pursuit of a systematic and algorithmic approach to investment has drawn considerable attention among traders and investment firms. Consequently, an effective computational method for evaluating potential risk factors and returns is crucial for the development of algorithmic trading strategies. In traditional finance and financial engineering research, statistical approaches have been widely applied to quantitative analysis. Meanwhile, investor demand for quantitative hedge funds has surged worldwide. In the current study, the multiperiod portfolio selection problem was considered in terms of the realistic transaction cost model, which is a major concern for quantitative hedge fund managers. We developed a dedicated multiagent-based deep reinforcement learning framework with a two-level nested agent structure to determine effective portfolio management methods with different objectives. In addition, we proposed a specially-designed reward function for investment performance evaluation and a novel policy network structure for trading decision-making. To efficiently identify specific asset attributes in a portfolio, each agent is equipped with a refined deep policy network and a special training method that enables the proposed reinforcement learning agent to learn risk transfer behaviors. The results revealed the effectiveness of our proposed framework, which outperformed several established or representative portfolio selection strategies.
查看更多>>摘要：Fault detection in non-stationary processes is a timely research topic in industrial systems. The conventional approaches based on principal component regression (PCR) and partial least-squares (PLS) cannot be loosely used for non-stationary and non-linear processes when the statistical behaviour of the measurements does not follow a Gaussian distribution with constant mean and standard deviation values. This paper introduces a new application of deep learning (DL), specifically a combination of proposed correlative stacked auto-encoder (C-SAE) and correlative deep neural networks (C-DNN) for output-related anomaly detection without complete decomposition of process variables with respect to quality output(s). With this aim, two new constructive and demoting loss functions are proposed to relatively decompose the process measurements with respect to their relevance to quality output variable(s). The loss functions are modified with the incorporation of non-linear correlation analysis and hence integrated into SAE and DNN structures to suggest a correlative SAE and DNN. The proposed C-SAE and C-DNN are integrated into a scheme with an inverted pyramid structure that enables output-related fault detection without limiting stationarity assumptions. Moreover, the proposed framework can be freely applied to both linear and non-linear processes. The performance of the proposed DL strategy is tested and validated on a non-stationary numerical example and Tennessee Eastman Process. The comparison results with recent approaches indicate the outperformance of the proposed approach for process output-related fault detection purposes.
查看更多>>摘要：Service composition and optimal selection (SCOS) plays a crucial role in cloud manufacturing (CMfg). While the existing service composition methods are hard to address the changes and uncertainties of CMfg dynamic environment. Therefore, a variable-length encoding genetic algorithm for structure-varying incremental service composition (ISC-GA) is proposed in this paper. Specifically, a novel variable-length encoding scheme containing structural information is proposed to describe the uncertain and changing process model. And the improved crossover and mutation algorithm suitable for individuals with nonlinear varying structure and incremental service composition is designed. It is realized by optimizing both the process structure and service instance combinations, and overcomes the drawbacks resulted from single preset process structure. Due to the difficulty of fitness computation caused by uncertain process structures, novelty is introduced as a new evolutionary pressure, and a novel framework for ISC-GA is presented, which helps to find both novel and high-performance solutions. Experimental results indicate the effectiveness of the proposed approach.
查看更多>>摘要：Learning unlabeled samples without deteriorating performance is a challenge in semi-supervised learning. In this paper, we propose a safe intuitionistic fuzzy twin support vector machine (SIFTSVM) for semi-supervised learning. In our SIFTSVM, whether an unlabeled sample should be learned by a twin support vector machine is determined by its plane intuitionistic fuzzy number. The unlabeled samples are learned gradually according to the current decision environment, which is safer and more precise than learning all of the unlabeled samples simultaneously. Interestingly, the iterative algorithm of our SIFTSVM obtains a solution to a mixed integer programming problem whose global solution corresponds to a classifier by learning the unlabeled samples with implicit labels. Experimental results on several synthetic datasets confirm the safety of our SIFTSVM for learning unlabeled samples, and the results on 56 groups of benchmark datasets demonstrate that our SIFTSVM outperforms the state-of-the-art semi-supervised classifiers on most groups.
查看更多>>摘要：Neural networks for classification aim at identifying the class label of new observation based training data containing instances whose category memberships are known. Therefore the data fed into neural networks has to be preprocessed to enhance its quality resulting in promoting the extraction of meaningful insights of data. However, the fact of processing data until you have the required high quality is challenging and time-consuming to manually search for the best method in a sequence of preprocessing independent methods. For feature scaling methods, they consist of scaling the dataset into the same range of data without monitoring data outliers that should eventually occur in the data source. Zscore for outlier's detection suffers from the issue of predefining the parameters. This paper discussed various approaches that are applied to scale features and detect outliers during data pre-processing. Thereafter, the paper proposed the algorithm that combines Zscore as an outlier's detection method with every classical feature scaling method in high-dimensional data. The proposed algorithm has benefits in selecting the optimal subset of methods from a sequence of chained methods, detecting outliers, and removing zero variance predictors. The study findings from five sample sizes revealed that the proposed method significantly excels the classical method in terms of accuracy. The outstanding from them was performed at the rate of 99.67% and had a significant difference of 0.20% over classical feature scaling.
查看更多>>摘要：The sustainable operation of sensor nodes in Wireless Sensor Network depends on the nodes’ adaptability with the environment. A sensor node strives to live longer using periodic sleep/awake activity. But it fails to achieve considerable success due to the node's inability to make the sleep/awake strategy adaptive to the environment. To this end, we propose an algorithm, ‘FL-Sleep’ which makes every node in the network to observe the ambient temperature and status of their parameters after every round of operation. Depending on their perception of the parameters, the nodes execute a sleep-scheduling strategy in the subsequent round. It makes the node evaluate its current state and decide the required action (’Active’, ‘Listen’ or ‘Sleep’) to perform. A node working in a favorable condition would decide the action with an optimistic attitude towards the parameters. In contrast, a critical condition of a node compels it to decide pessimistically. This qualitative measurement provides a precise understanding of the environment. ‘FL-Sleep’ works on hesitant fuzzy logic-based Multi-Criteria Decision Making method and is found to improve the network's lifetime by 247.11% compared to BMAC, by 68.56% compared to SOPC, and by 77.2% compared to RL-Sleep. The best lifetime of nodes is obtained when the network is organized in spiral topology. ‘FL-Sleep’ shows better performance in terms of packet-delivery-ratio, energy efficiency, and the number of active nodes in the network compared to BMAC, SOPC, and RL-Sleep.
NSTL
Elsevier