Solving the stochastic support vector regression with probabilistic constraints by a high-performance neural network model

This paper offers a recurrent neural network to support vector machine (SVM) learning in stochastic support vector regression with probabilistic constraints. The SVM is first converted into an equivalent quadratic programming (QP) formulation in linear and nonlinear cases. An artificial neural network for SVM learning is then proposed. The presented neural network framework guarantees obtaining the optimal solution of the SVM problem. The existence and convergence of the trajectories of the network are studied. The Lyapunov stability for the considered neural network is also shown. The efficiency of the proposed method is shown by three illustrative examples.

     

The Fuzzy Logic Application in Volume Fractions Prediction of the Annular Three-Phase Flows

In this paper, the volume fractions in the annular three-phase flow are measured based on a dual energy metering system consisting of \(^{152}\)Eu and \(^{137}\)Cs and one NaI detector, and then modeled by fuzzy logic. Since the summation of volume fractions are constant (equal to 100%), therefore the fuzzy network must predict only two volume fractions. In this study, three fuzzy networks are applied. The first network is utilized to predict the gas and water volume fractions. The next one is applied to predict the gas and oil volume fractions, and the last one to predict the water and oil volume fractions. In the next step, the numerically obtained data from MCNP-X code, must be imported to the fuzzy models. Then, the average errors of these three networks are computed and compared. The network which has the least error is selected as the best predictor model. According to the modeling results, the best fuzzy network, predicts the gas and water volume fractions with the mean relative error of less than 0.3%, which shows that the fuzzy logic can predict the results precisely.     

Representing Local Dynamics of Water Resource Systems through a Data-Driven Emulation Approach

Water Resources Management - Water resource systems are under enormous pressures globally. To diagnose and quantify potential vulnerabilities, effective modeling tools are required to represent the...     

A self‐learning stream classifier for flow‐based botnet detection

Botnets have been recently recognized as one of the most formidable threats on the Internet. Different approaches have been designed to detect these types of attacks. However, as botnets evolve their behavior to mislead the signature‐based detection systems, learning‐based methods may be deployed to provide a generalization capacity in identifying unknown botnets. Developing an adaptable botnet detection system, which incrementally evolves with the incoming flow stream, remains as a challenge. In this paper, a self‐learning botnet detection system is proposed, which uses an adaptable classification model. The system uses an ensemble classifier and, in order to enhance its generalization capacity, updates its model continuously on receiving new unlabeled traffic flows. The system is evaluated with a comprehensive data set, which contains a wide variety of botnets. The experiments demonstrate that the proposed system can successfully adapt in a dynamic environment where new botnet types are observed during the system operation. We also compare the system performance with other methods.     

Solving general convex nonlinear optimization problems by an efficient neurodynamic model

In this paper, a neural network model is constructed on the basis of the duality theory, optimization theory, convex analysis theory, Lyapunov stability theory and LaSalle invariance principle to solve general convex nonlinear programming (GCNLP) problems. Based on the Saddle point theorem, the equilibrium point of the proposed neural network is proved to be equivalent to the optimal solution of the GCNLP problem. By employing Lyapunov function approach, it is also shown that the proposed neural network model is stable in the sense of Lyapunov and it is globally convergent to an exact optimal solution of the original problem. The simulation results also show that the proposed neural network is feasible and efficient.     

Numerical investigation of the effects of combining sewer junction characteristics on the hydraulic parameters of flow in fully surcharged condition

In this study, the effects of different parameters on flow structure and flow hydraulics for combining sewer junction consisting of manhole and lateral inflow were investigated numerically. For this purpose, the effects of dimension variations, main and lateral pipe slope, angle and the joint position of lateral pipe beside manhole, manhole geometry, and the outflow rate of manhole on flow structure and energy loss coefficients were investigated. Results show that the rate of energy loss decreases by heightening the installation position of lateral pipe beside manhole. In addition, it was shown that the slope variations of main and lateral pipes and outflow rate do not affect the rate of energy loss.     

Material bead deposition with 2 + 2 ½ multi-axis machining process planning strategies with virtual verification for extruded geometry

Process planning for hybrid manufacturing, where additive operations can be interlaced with machining operations, is in its infancy. New plastic- and metal-based hybrid manufacturing systems are being developed that integrate both additive manufacturing (AM) and subtractive (machining) operations. This introduces new process planning challenges. The focus of this research is to explore process planning solution approaches when using a hybrid manufacturing approach. Concepts such as localized AM build ups, adding stock to a CAD model or section for subsequent removal, and machining an AM stock model are investigated and illustrated using virtual simulations. A case study using a hybrid laser cladding process is used to demonstrate the opportunities associated with a hybrid solution. However, unlike machining, the process characteristics from system to system vary greatly. These are portrayed via a high power, high material deposition feed rate laser cladding system. There are unique challenges associated with AM processes and hybrid manufacturing. New tools and design rules need to be developed for this manufacturing solution to reach its potential.     

An application of a merit function for solving convex programming problems

This paper presents a gradient neural network model for solving convex nonlinear programming (CNP) problems. The main idea is to convert the CNP problem into an equivalent unconstrained minimization problem with objective energy function. A gradient model is then defined directly using the derivatives of the energy function. It is also shown that the proposed neural network is stable in the sense of Lyapunov and can converge to an exact optimal solution of the original problem. It is also found that a larger scaling factor leads to a better convergence rate of the trajectory. The validity and transient behavior of the neural network are demonstrated by using various examples.     

Artificial photosynthesis at soft interfaces

The concept of artificial photosynthesis at a polarised liquid membrane is presented. It includes two photosystems, one at each interface for the hydrogen and oxygen evolution respectively. Both reactions involve proton coupled electron transfer reactions, and some ultrafast steps at the photosensitization stage.     

ERP: a literature survey

Most of organizations have a “functional structure,” which is composed of various functional units. In some cases, each functional unit works towards their own goals and objectives, rather than the organizational goals. This is further aggravated because information flow is restricted by functions, and even when other functional units want to take a systemic view, they do not have the needed information to do so. These are the issues addressed by “Enterprise Resource Planning” (ERP) software solutions providing a common and consistent system to capture information organization-wide, with minimum redundancy. Today, many organizations acquire and implement ERP to improve their operational performance and create strategic value; however, they fail to achieve these objective due to lack of knowledge and better understanding of ERP and its lifecycle. Although, so far, hundreds of research article are published separately focusing on ERP and various issues related to its lifecycle and management, there is no survey and overview of the article published in different top-tier journals. We believe that a survey of the articles related to ERP provides better understanding of ERP and attempts to create an information bank of the published articles these in turn, contribute to improving the performance of an enterprise in terms of achieving its strategic value creation goals. In this study, we attempt to summarize our survey and review of the articles related to ERP.