Accurate streamflow (SF) forecasting is crucial for effective water-resource management amid global climate change. Traditional ensemble SF-forecasting methods, relying on historical data and watershed characteristics, often produce uncertainties in their input, structure, and parameters, reducing their forecasting accuracy. This study introduces a variance-combined bias-correction (VCB) method, integrated with Internet of Things (IoT) technology to improve ensemble SF forecasts’ accuracy and responsiveness. The VCB method significantly improves the SF-forecasting performance by incorporating variance information from ensemble forecasts along with the ensemble mean. We apply the method to the Shiquan Reservoir in China’s Han River basin, and the results show that the VCB method outperforms the Bayesian joint probability (BJP) method, achieving an increases of 8.8% in the Nash-Sutcliffe efficiency (NSE), 0.7% in the Pearson correlation coefficient (PCC), 2.1% in the qualified rate (QR), and a 7.2% reduction in the mean absolute percentage error (MAPE). Furthermore, IoT technology integration improves method inputs’ accuracy and timeliness, showing the strongest performance during extreme weather events. Thus, by improving uncertainty management and forecasting accuracy, the IoT-integrated VCB method provides more effective support for water-resource management. Future research should apply this approach to diverse hydrological contexts and explore deeper integration with machine-learning techniques.

Summary Feature extraction is a fundamental step in the feature matching task. A lot of studies are devoted to feature extraction. Recent researches propose to extract features by pre‐trained neural networks, and the output is used for feature matching. However, the quality and the quantity of the features extracted by these methods are difficult to meet the requirements for the practical applications. In this article, we propose a two‐stage object‐aware‐based feature matching method. Specifically, the proposed object‐aware block predicts a weighted feature map through a mask predictor and a prefeature extractor, so that the subsequent feature extractor pays more attention to the key regions by using the weighted feature map. In addition, we introduce a state‐of‐the‐art model estimation algorithm to align image pair as the input of the object‐aware block. Furthermore, our method also employs an advanced outlier removal algorithm to further improve matching quality. Experimental results show that our object‐aware‐based feature matching method improves the performance of feature matching compared with several state‐of‐the‐art methods.

Multi-view data often contain redundant information that cannot be simply spliced. Many existing methods for processing them by assigning weights to each view cannot capture features dynamically. Therefore, we propose a multi-view deep matrix factorization method via neural networks that captures semantic hierarchical information of the data and dynamically produces a consistent representation using the complementarity of multi-view features. Due to the usefulness of deep matrix factorization, the generated representation is easily interpretable. The proposed method yields a harmonized representation directly from multi-view data without an extra weight learning process. In addition, we use a multi-path network to search for a consensual solution and obtain an optimal result. Additional feature optimization is used to enhance the discriminative characterization of the representation matrix. Finally, experiments on four real-world datasets show that the proposed method is superior to state-of-the-arts.

Recent researches on multiview learning have received widespread attention due to the increasing generalization of multiview data. As an effective probabilistic model, random walk has also shown encouraging performance in various fields. To further exploit the potential of utilizing random walk schemes to address multiview learning problems, this article proposes a simple yet efficient multigraph random walk scheme for both multiview clustering and semisupervised classification tasks. The proposed model integrates random walk with multiview learning, and recursively learns a globally stable probability distribution matrix from multiple views, on the basis of which the label indicator is obtained in the scene of clustering or semisupervised classification. Furthermore, an adaptive weight vector is learned to incorporate the diversity and complementarity of multiview data. Besides, the relationships between the proposed scheme and spectral clustering, neighborhood embedding and manifold embedding are analyzed theoretically. Finally, comprehensive comparative experiments are conducted with several state-of-the-art multiview clustering and semisupervised classification methods on eight real-world datasets. The experimental results demonstrate the superiority of the proposed method in terms of both clustering and classification performance.

As a fundamental and critical task in feature-based remote sensing image registration, feature matching refers to establishing reliable point correspondences from two images of the same scene. In this article, we propose a simple yet efficient method termed linear adaptive filtering (LAF) for both rigid and nonrigid feature matching of remote sensing images and apply it to the image registration task. Our algorithm starts with establishing putative feature correspondences based on local descriptors and then focuses on removing outliers using geometrical consistency priori together with filtering and denoising theory. Specifically, we first grid the correspondence space into several nonoverlapping cells and calculate a typical motion vector for each one. Subsequently, we remove false matches by checking the consistency between each putative match and the typical motion vector in the corresponding cell, which is achieved by a Gaussian kernel convolution operation. By refining the typical motion vector in an iterative manner, we further introduce a progressive strategy based on the coarse-to-fine theory to promote the matching accuracy gradually. In addition, an adaptive parameter setting strategy and posterior probability estimation based on the expectation-maximization algorithm enhance the robustness of our method to different data. Most importantly, our method is quite efficient where the gridding strategy enables it to achieve linear time complexity. Consequently, some sparse point-based tasks may inspire from our method when they are achieved by deep learning techniques. Extensive feature matching and image registration experiments on several remote sensing data sets demonstrate the superiority of our approach over the state of the art.

Image cosegmentation is a newly emerging research area in image processing. It refers to the problem of segmenting the common objects simultaneously in multiple images by utilizing the similarity of foreground regions among these images. In this paper, a new active contour model is proposed by using shape-similarity and foreground discovery scheme. The foreground discovery scheme is used to obtain the rough contours of the common objects which are used as initial evolution curves. The energy function of the proposed model includes two parts: an intra-image energy and an inter-image energy. The intra-image energy explores the differences between foreground regions and background regions in each image. And the inter-image energy is used to explore the similarities of the common objects among target images, which composes of a region color feature energy term and a shape constraint energy term. The region color feature term indicates the foreground consistency and the background consistency among the images; and the shape constraint energy term allows the global changes of shapes and truncates the local variation caused by misleading features. Experimental results show that the proposed model can improve the accuracy of the image cosegmentation significantly through regularizing the changes of shapes.

E-commerce websites such as Yelp.com, allow users to write online reviews of products and services, so new customers can have quick access to user experiences, covering everything from auto-repair to hospitals. However, a typical user may find it difficult to identify a topic of interest due to the overwhelming amount of review information. To deal with this issue, the Latent Dirichlet Allocation (LDA) model can be used to associate meaningful terms with text-based reviews, permitting keyword retrieval of individual documents. LDA is a powerful unsupervised learning approach, which has been widely used for topic modeling as well as in other related fields. A conventional implementation of LDA is through the Markov-Chain Monte-Carlo methodology, called Collapsed Gibbs Sampling (CGS). However, due to the usage of random numbers in the CGS approach, results from multiple trials on the same data are usually inconsistent. To avoid this tendency, we revise the conventional LDA approach using Variational Gibbs Sampling (VGS). VGS eliminates random numbers, and thus leads to consistent results as well as better performance. Our case study shows that our improved LDA can be used to automatically identify keywords and topics in online hospital reviews. Due to the usage of VGS, the accuracy of topic identification has been consistently improved.

Conventional single-vector-based direct torque control (SVB-DTC) has been widely researched owing to its simple control structure, parameter robustness and fast response. However, the SVB-DTC often features large torque and flux ripple. To suppress the large ripple in torque and stator flux, recently, a duty-cycle-based DTC (DCB-DTC) scheme applying a zero vector and a nonzero vector to the drive during every control period has been presented. However, the torque and flux ripple fail to be reduced to a minimal value owing to the fixed vector combination of a zero and a nonzero vector. A generalized two-vector-based DTC (GTV-DTC) for permanent-magnet synchronous machine (PMSM) drives which relaxes the vector combination to two arbitrary voltage vectors is proposed. In contrast to the DCB-DTC, two nonzero vectors might be chosen in the proposed method. Then the torque ripple can be reduced by the broad possible vector combination. In addition, a simple analytical and graphical method is proposed to select two voltage vectors and to determine their optimal durations based on the principle of stator flux vector error minimization. The simulation results are provided to verify the outstanding performance of the proposed method that, compared to prior SVB-DTC and DCB-DTC, the proposed GTV-DTC achieves much better steady-state performance and similar transient-state performance without significantly increasing computational burden.

Due to large amounts of reviews for many similar online products, users often feel difficult to determine which products have the most desirable features that they want.In this paper, we propose a model-based approach to analyzing online reviews and identifying the strengths and weaknesses of a product by its product features.We propose a Sentence Level Topic Model (SLTM), which can classify review sentences into different classes corresponding to different product features.The model contains a hidden layer, called the topic layer, between corpus and words.Once a SLTM has been trained with sufficient labeled data points, it can identify the most related topic (i.e., product feature) for each sentence.To capture a reviewer' opinion, we perform sentiment analysis for each review sentence, and derive the weighted feature preference vectors for the review.Finally, we combine the results of all review comments for a product into a review summary.The case study shows that by comparing the review summaries of similar online products, users may have a much easier time to find their desired products.

The reliability and availability of a cloud-based system play an important role in evaluating its system performance.Due to the promised high reliability of physical facilities provided for cloud services, software faults have become a major factor for failures of cloud-based systems.In this paper, we focus on the software aging phenomenon where system performance may be progressively degraded due to exhaustion of system resources, fragmentation and accumulation of errors.We present a proactive technique, called software rejuvenation, to counteract the software aging problem.The dynamic fault tree (DFT) formalism is adopted to model the system reliability before and during a software rejuvenation process in an aging cloud-based system.Then it is converted into Markov Chains to derive the system reliability function.We use a case study of a cloud-based system to illustrate the validity of our approach.Based on the reliability analysis results, we show how to estimate software rejuvenation schedules that can keep the system reliability above a predefined critical level for required system availability.

Despite the many advantages provided by cloud-based storage services, there are still major concerns such as security, reliability and confidentiality of data stored in the cloud.In this paper, we propose a reliable and secure cloud storage schema using multiple service providers.Different from existing approaches to achieving data reliability using redundancy at the server side, we propose a reliable and secure cloud storage schema that can be implemented at the client side.In our approach, we view multiple cloud-based storage services as virtual independent disks for storing redundant data encoded using erasure codes.Since each independent cloud service provider has no access to a user's complete data, the data stored in the cloud would not be easily compromised.Furthermore, the failure or disconnection of a service provider will not result in the loss of a user's data as the missing data pieces can be readily recovered.To demonstrate the feasibility of our approach, we developed a prototype cloud-based storage system that breaks a data file into multiple data pieces, generates an optimal number of checksum pieces, and uploads them into multiple cloud storages.Upon the failure of a cloud storage service, the application can quickly restore the original data file from the available pieces of data.The experimental results show that our approach is not only secure and fault-tolerant, but also very efficient due to concurrent data processing.

Correctly measuring the reliability and availability of a cloud-based system is critical for evaluating its system performance. Due to the promised high reliability of physical facilities provided for cloud services, software faults have become one of the major factors for the failures of cloud-based systems. In this paper, we focus on the software aging phenomenon where system performance may be progressively degraded due to exhaustion of system resources, fragmentation and accumulation of errors. We use a proactive technique, called software rejuvenation, to counteract the software aging problem. The dynamic fault tree (DFT) formalism is adopted to model the system reliability before and during a software rejuvenation process in an aging cloud-based system. A novel analytical approach is presented to derive the reliability function of a cloud-based Hot SPare (HSP) gate, which is further verified using Continuous Time Markov Chains (CTMC) for its correctness. We use a case study of a cloud-based system to illustrate the validity of our approach. Based on the reliability analytical results, we show how cost-effective software rejuvenation schedules can be created to keep the system reliability consistently staying above a predefined critical level.

Software Engineering (SE) and Knowledge Engineering (KE) are closely related disciplines with goals of turning the development process of software systems and knowledge-based systems, respectively, into engineering disciplines. In particular, they together can provide systematic approaches for engineering intelligent software systems more efficiently and cost-effectively. As there is a large overlap between the two disciplines, the interplay is vital for both to be successful. In this paper, we divide the intersection of SE and KE into three subareas, namely Knowledge-Supported Software Engineering (KSSE), Engineering Knowledge as a Software (EKaaS), and Intelligent Software System Engineering (ISSE). For each subarea, we describe the challenges along with the current trends, and predict the future research directions that may have the most potential for success.

Despite the popularity and many advantages of using cloud data storage, there are still major concerns about the data stored in the cloud, such as security, reliability and confidentiality. In this paper, we propose a reliable and secure distributed cloud data storage schema using Reed-Solomon codes. Different from existing approaches to achieving data reliability with redundancy at the server side, our proposed mechanism relies on multiple cloud service providers (CSP), and protects users’ cloud data from the client side. In our approach, we view multiple cloud-based storage services as virtual independent disks for storing redundant data encoded with erasure codes. Since each CSP has no access to a user’s complete data, the data stored in the cloud would not be easily compromised. Furthermore, the failure or disconnection of a CSP will not result in the loss of a user’s data as the missing data pieces can be readily recovered. To demonstrate the feasibility of our approach, we developed a prototype distributed cloud data storage application using three major CSPs. The experimental results show that, besides the reliability and security related benefits of our approach, the application outperforms each individual CSP for uploading and downloading files.

PDE based image segmentation methods are the state-of-the-art methods due to the high accuracy and continuity of detected edges. Some examples are the region-scalable fitting (RSF) model and the local image fitting energy (LIF) model. But they suffer from the high computing complexity and instability. Recently, the globally convex method and the split Bregman method are introduced to overcome this problem. In this paper, a globally convex version of the LIF model is proposed and then the split Bregman method is used to solve the model. Experiments show that this model is more efficient than the RSF model and the LIF model while with similar segmentation results.

With the significant increase of available item listings in popular online auction houses nowadays, it becomes nearly impossible to manually investigate the large amount of auctions and bidders for shill bidding activities, which are a major type of auction fraud in online auctions. Automated mechanisms such as data mining techniques were proved to be necessary to process this type of increasing workload. In this paper, we first present a framework of Real-Time Self-Adaptive Classifier (RT-SAC) for identifying suspicious bidders in online auctions using an incremental neural network approach. Then, we introduce a clustering module that characterizes bidder behaviors in measurable attributes and uses a hierarchical clustering mechanism to create training datasets. The neural network in RT-SAC is initialized with the training datasets, which consist of labeled historical auction data. Once initialized, the network can be trained incrementally to gradually adapt to new bidding data in real time, and thus, it supports efficient detection of suspicious bidders in online auctions. Finally, we utilize a case study to demonstrate how parameters in RT-SAC can be tuned for optimal operations and how our approach can be used to effectively identify suspicious online bidders in real time.

Cloud computing allows for access to ubiquitous data storage and powerful computing resources through the use of web services. There are major concerns, however, with data security, reliability, and availability in the cloud. In this paper, we address these concerns by introducing a novel security mechanism for secure and fault-tolerant cloud information storage. The information storage model follows the RAID (Redundant Array of Independent Disks) concept by considering cloud service providers as independent virtual disk drives. As such, the model utilizes multiple cloud service providers as a cloud cluster for information storage, and a service directory for management of the cloud clusters including service query, key management, and cluster restoration. Our approach not only supports maintaining the confidentiality of the stored data, but also ensures that the failure or compromise of an individual cloud provider in a cloud cluster will not result in a compromise of the overall data set. To ensure a correct design, we present a formal model of the security mechanism using hierarchical colored Petri nets (HCPN), and verify some key properties of the model using model checking techniques.

This paper describes the design of a decision support system for shill detection in online auctions. To assist decision making, each bidder is associated with a type of certification, namely shill, shill suspect, or trusted bidder, at the end of each auction's bidding cycle. The certification level is determined on the basis of a bidder's bidding behaviors including shilling behaviors and normal bidding behaviors, and thus fraudulent bidders can be identified. In this paper, we focus on representing knowledge about bidders from different aspects in online auctions, and reasoning on bidders' trustworthiness under uncertainties using Dempster–Shafer theory of evidence. To demonstrate the feasibility of our approach, we provide a case study using real auction data from eBay. The analysis results show that our approach can be used to detect shills effectively and efficiently. By applying Dempster–Shafer theory to combine multiple sources of evidence for shill detection, the proposed approach can significantly reduce the number of false positive results in comparison to approaches using a single source of evidence.

Computer system reliability is conventionally modeled and analyzed using techniques such as fault tree analysis and reliability block diagrams (RBDs), which provide static representations of system reliability properties. A recent extension to RBDs, called dynamic RBDs (DRBD), defines a framework for modeling the dynamic reliability behavior of computer-based systems. However, analyzing a DRBD model in order to locate and identify design errors, such as a deadlock error or faulty state, is not trivial when done manually. A feasible approach to verifying it is to develop its formal model and then analyze it using programmatic methods. In this paper, we first define a reliability markup language that can be used to formally describe DRBD models. Then, we present an algorithm that automatically converts a DRBD model into a colored Petri net. We use a case study to illustrate the effectiveness of our approach and demonstrate how system properties of a DRBD model can be verified using an existing Petri net tool. Our formal modeling approach is compositional; thus, it provides a potential solution to automated verification of DRBD models.

Security modeling for agents has been one of the most challenging issues in developing practical mobile agent software systems. In the past, researchers have developed mobile agent systems with emphasis either on protecting mobile agents from malicious hosts or protecting hosts from malicious agents. In this paper, we propose a security based mobile agent system architecture that provides a general solution to protecting both mobile agents and agent hosts in terms of agent communication and agent migration. We present a facilitator agent model that serves as a middleware for secure agent communication and agent migration. The facilitator agent model, as well as the mobile agent model, is based on agent-oriented G-nets — a high level Petri net formalism. To illustrate our formal modeling technique for mobile agent systems, we provide an example of agent migration to show how a design error can be detected.

Agents are becoming one of the most important topics in distributed and autonomous decentralized systems, and there are increasing attempts to use agent technologies to develop large-scale commercial and industrial software systems. The complexity of such systems suggests a pressing need for system modeling techniques to support reliable, maintainable, and extensible design. G-nets are a type of Petri net defined to support system modeling in terms of a set of independent and loosely-coupled modules. In this paper, we customize the basic G-net model to define a so-called "agent-based G-net" that can serve as a generic model for agent design. Then, to progress from an agent-based design model to an agent-oriented model, new mechanisms to support inheritance modeling are introduced. To illustrate our formal modeling technique for multiagent systems, an example of an agent family in electronic commerce is provided. Finally, we demonstrate how we can use model checking to verify some key behavioral properties of our agent model. This is facilitated by the use of an existing Petri net tool.

With the increasing importance of complex software systems in the software industry, the need for using agent technologies to develop large-scale commercial and industrial software systems is growing rapidly. Such systems are complex, and there is a pressing need for system modeling techniques to support reliable, maintainable and extensible design. G-nets are a type of Petri net defined to support the modeling of a system as a set of independent and loosely-coupled modules. In this paper, we first introduce an extension of G-nets - the agent-based G-net - as a generic model for agent design. Then, to progress from an agent-based design model to an agent-oriented model, new mechanisms to support inheritance modeling are introduced. To illustrate our formal modeling technique for multi-agent systems, an example of an agent family in electronic commerce is provided.

Agents are becoming one of the most important topics in distributed and autonomous decentralized systems (ADS), and there are increasing attempts to use agent technologies to develop software systems in electronic commerce. Such systems are complex and there is a pressing need for system modeling techniques to support reliable, maintainable and extensible design. G-Nets are a type of Petri net defined to support modeling of a system as a set of independent and loosely-coupled modules. The authors first introduce an extension of G-Net, agent based G-Net, as a generic model for agent design. Then new communication mechanisms are introduced to support asynchronous message passing among agents. To illustrate that our formal modeling technique is effective for agent modeling in electronic commerce, a price-negotiation protocol example between buyers and sellers is provided. Finally, by analyzing an ordinary Petri net reduced from our agent based G-Net models, we conclude that our agent based G-Net models are L3-live, concurrent and effective for agent communications.

G-nets are a type of Petri net defined to support the modeling of a system as a set of independent and loosely-coupled modules. The modular features of G-nets provide support for incremental design and successive modification, however the G-net formalism is not fully object-oriented due to a lack of support for inheritance. We introduce extensions to G-nets to support explicit modeling of inheritance. Bounded buffer examples are used, which we define as subclasses of an unbounded buffer, to illustrate the expressive power of the extended G-net models. Various forms of inheritance are formalized and discussed in the context of concurrent object-oriented design. In addition, the inheritance anomaly problem is examined and discussed.