One-class classification problem has been investigated thoroughly for past decades. Among one of the most effective neural network approaches for one-class classification, autoencoder has been successfully applied for...
详细信息
One-class classification problem has been investigated thoroughly for past decades. Among one of the most effective neural network approaches for one-class classification, autoencoder has been successfully applied for many applications. However, this classifier relies on traditional learning algorithms such as backpropagation to train the network, which is quite time-consuming. To tackle the slow learning speed in autoencoder neural network, we propose a simple and efficient one-class classifier based on extreme learning machine (ELM). The essence of ELM is that the hidden layer need not be tuned and the output weights can be analytically determined, which leads to much faster learning speed. The experimental evaluation conducted on several real-world benchmarks shows that the ELM based one-class classifier can learn hundreds of times faster than autoencoder and it is competitive over a variety of one-class classification methods.
The problem of developing an control algorithm for a queuing system is considered. This system has a finite number of states the dynamics of which is described by a conditional Markov chain;the system is observed usin...
详细信息
The problem of developing an control algorithm for a queuing system is considered. This system has a finite number of states the dynamics of which is described by a conditional Markov chain;the system is observed using indicators whose readings are error prone. Optimal and approximately optimal solutions based on the theory of systems with random jump structure are found. By way of example, the problem of synthesis of an approximately optimal algorithm for the recognition of state and for control of aviation raids on a military facility that is alternatively damaged and restored in the course of air combat operations.
Imperfect preventive maintenance (PM) activities are very common in industrial systems. For condition-based maintenance (CBM), it is necessary to model the failure likelihood of systems subject to imperfect PM activit...
详细信息
Imperfect preventive maintenance (PM) activities are very common in industrial systems. For condition-based maintenance (CBM), it is necessary to model the failure likelihood of systems subject to imperfect PM activities. In this paper, the models in the field of survival analysis are introduced into CBM. Namely, the generalized accelerated failure time (AFT) frailty model is investigated to model the failure likelihood of industrial systems. Further, on the basis of the traditional maximum likelihood (ML) estimation and expectation maximization (EM) algorithm, the hybrid ML-EM algorithm is investigated for the estimation of parameters. The hybrid iterative estimation procedure is analyzed in detail. In the evaluation experiment, the generated data of a typical degradation model are verified to be appropriate for the real industrial processes with imperfect PM activities. The estimates of the model parameters are calculated using the training data. Then, the performance of the model is analyzed through the prediction of remaining useful life (RUL) using the testing data. Finally, comparison between the results of the proposed model and the existing model verifies the effectiveness of the generalized AFT frailty model.
This study proposes an algorithm for generating the associated Boolean expression in VHDL, given a ladder diagram (LD) as the input. The purpose of the algorithm is to implement of field-programmable gate array-(FPGA-...
详细信息
This study proposes an algorithm for generating the associated Boolean expression in VHDL, given a ladder diagram (LD) as the input. The purpose of the algorithm is to implement of field-programmable gate array-(FPGA-) based programmable logic controllers (PLCs), where an effective conversion from an LD to its associated Boolean expressions seems rarely mentioned. Based on this core thought, the conversion process of the algorithm first involves abstracting and expressing the encountered LD as an activity-on-vertex (AOV) graph. Next, an AND-OR tree in which AND-nodes and OR-nodes connote the series and the parallel relationships between the vertices of the AOV graph is constructed based on the AOV graph. Therefore, by a traversal to the AND-OR tree, the associated Boolean expression, as the output of the algorithm, can be easily obtained in VHDL. The proposed algorithm is then verified with an illustrative example, wherein a complicated LD is given as the input.
The nonlocal means algorithm is widely used in image denoising, but this algorithm does not work well for high-intensity noise. To overcome this shortcoming, we establish a coupled iterative nonlocal means model in th...
详细信息
The nonlocal means algorithm is widely used in image denoising, but this algorithm does not work well for high-intensity noise. To overcome this shortcoming, we establish a coupled iterative nonlocal means model in this paper. Considering the computation complexity of the new model, we realize it by using multiscale wavelet transform and propose an asymptotic nonlocal filtering algorithm which can reduce the influence of noise on similarity estimation and computation complexity. Moreover, we build a new nonlocal weight function based on the structure similarity index. Simulation results indicate that the proposed approach cannot only remove the noise but also preserve the structure of image and has good visual effects, especially for highly degenerated images.
Smart homes (SHs) are crucial parts for demand response management (DRM) of smart grid (SG). The aim of SHs based demand response (DR) is to provide a flexible two-way energy feedback whilst (or shortly after) the con...
详细信息
Smart homes (SHs) are crucial parts for demand response management (DRM) of smart grid (SG). The aim of SHs based demand response (DR) is to provide a flexible two-way energy feedback whilst (or shortly after) the consumption occurs. It can potentially persuade end-users to achieve energy saving and cooperate with the electricity producer or supplier to maintain balance between the electricity supply and demand through the method of peak shaving and valley filling. However, existing solutions are challenged by the lack of consideration between the wide application of fiber power cable to the home (FPCTTH) and related users' behaviors. Based on the new network infrastructure, the design and development of smart DR systems based on SHs are related with not only functionalities as security, convenience, and comfort, but also energy savings. A new multirouting protocol based on Kruskal's algorithm is designed for the reliability and safety of the SHs distribution network. The benefits of FPCTTH-based SHs are summarized at the end of the paper.
The large number of potential applications for robotic odor source localization has motivated the development of a variety of plume tracking algorithms, the majority of which work in restricted two-dimensional scenari...
详细信息
ISBN:
(纸本)9781509037636
The large number of potential applications for robotic odor source localization has motivated the development of a variety of plume tracking algorithms, the majority of which work in restricted two-dimensional scenarios. In this paper, we introduce a distributed algorithm for 3-D plume tracking using a system of ground and aerial robots in formation. We propose an algorithm that takes advantage of spatially distributed measurements to track the plume in 3-D and lead the robots to the source by integrating three behaviors - upwind movement, plume centering, and Laplacian feedback formation control. We evaluate this strategy in simulation and with real robots in a wind tunnel. For a source close to the ground, results show that a team of robots running our algorithm reaches the source with low lateral error while also tracing the horizontal and vertical plume shape.
Text plays an important role in daily life because of its rich information, thus automatic text detection in natural scenes has many attractive applications. However, detecting and recognising such text is always a ch...
详细信息
Text plays an important role in daily life because of its rich information, thus automatic text detection in natural scenes has many attractive applications. However, detecting and recognising such text is always a challenging problem. In this study, the authors propose a method which extends the widely-used stroke width transform by two steps of edge analysis, namely candidate edge recombination and edge classification. A new method that recognises text through candidate edge recombination and candidate edge recognition is also proposed. In the step of candidate edge recombination, they use the idea of over-segmentation and region merging. To separate text edge from background, the edge of the input image is first divided into small segments. Then, neighbour edge segments are merged, if they have similar stroke width and colour. Through this step, each character is described by one candidate boundary. In the step of boundary classification, candidate boundaries are aggregated into text chains, followed by chain classification using character-based and chain-based features. To recognise text, the grey image is extracted based on the location of each candidate edge after the step of candidate edge recombination. Then, histogram of gradient features and a classifier are used to recognise each character. To evaluate the effectiveness of their method, the algorithm is run on the ICDAR competition dataset and Street View Text database. The experimental results show that the proposed method provides promising performance in comparison with the existing methods.
Based on a graph-theoretic concept of a cluster, dominant sets clustering has been shown to be an attractive clustering algorithm with many useful properties. In this study, the authors conduct a comprehensive study o...
详细信息
Based on a graph-theoretic concept of a cluster, dominant sets clustering has been shown to be an attractive clustering algorithm with many useful properties. In this study, the authors conduct a comprehensive study of related issues in dominant sets clustering, in an endeavour to explore the potential of this algorithm and obtain the best clustering results. Specifically, they empirically investigate how similarity parameters, similarity measures and game dynamics influence the dominant sets clustering results. From experiments on eight datasets, they conclude that distance-based similarity measures perform evidently better than cosine and histogram intersection similarity measures potentially, and they need to find the best-performing similarity parameter to make use of this advantage. They then study the effect of similarity parameter on dominant sets clustering results and induce the range of the best-performing similarity parameters. Furthermore, they find that the recently proposed infection and immunisation dynamics performs better than the replicator dynamics in most cases while being much more efficient than the latter. These observations are helpful in applying dominant sets clustering to practical problems, and also indicate directions for further improvement of this algorithm.
A set of explicit finite difference schemes with large stencil was optimized to obtain maximum resolution characteristics for various spatial truncation orders. The effect of integral interval range of the objective f...
详细信息
A set of explicit finite difference schemes with large stencil was optimized to obtain maximum resolution characteristics for various spatial truncation orders. The effect of integral interval range of the objective function on the optimized schemes' performance is discussed. An algorithm is developed for the automatic determination of this integral interval. Three types of objective functions in the optimization procedure are compared in detail, which show that Tam's objective function gets the best resolution in explicit centered finite difference scheme. Actual performances of the proposed optimized schemes are demonstrated by numerical simulation of three CAA benchmark problems. The effective accuracy, strengths, and weakness of these proposed schemes are then discussed. At the end, general conclusion on how to choose optimization objective function and optimization ranges is drawn. The results provide clear understanding of the relative effective accuracy of the various truncation orders, especially the trade-off when using large stencil with a relatively high truncation order.
暂无评论