In this paper a stochastic adaptive method has been developed to solve stochastic linear problems by a finite sequence of Monte-Carlo sampling estimators. The method is based on the adaptive regulation of the size of Monte-Carlo samples and a statistical termination procedure taking into consideration statistical modelling accuracy. Our approach distinguishes itself by the treatment of accuracy of the solution in a statistical manner, testing the hypothesis of optimality according to statistical criteria, and estimating confidence intervals of the objective and constraint functions. To avoid “jamming” or “zigzagging” solving a constraint problem we implement the ε–feasible direction approach. The proposed adjustment of a sample size, when it is taken inversely proportional to the square of the norm of the Monte-Carlo estimate of the gradient, guarantees convergence a. s. at a linear rate. The numerical study and examples in practice corroborate theoretical conclusions and show that the developed procedures make it possible to solve stochastic problems with sufficient accuracy by the means of an acceptable size of computations.
The paper proposes a technology for mass optimization of two-dimensional body applying genetic algorithms. Main attention is focused on geometry of 2D body, i. e. search for optimal coordinates of body points. Direct analysis of 2D body – von Mises stress determination – is performed using original program based on finite element method. The set of design parameters contains the coordinates of body points in 2D space. The results of numerical experiments proved the proposed technology to be efficient tool for solution of 2D body mass optimization problem.
Cloud computing can be defined as a new style of computing in which dynamically scala-ble and often virtualized resources are provided as a services over the Internet. Advantages of the cloud computing technology include cost savings, high availability, and easy scalability. Voas and Zhang adapted six phases of computing paradigms, from dummy termi-nals/mainframes, to PCs, networking computing, to grid and cloud computing. There are four types of cloud computing: public cloud, private cloud, hybrid cloud and community. The most common and well-known deployment model is Public Cloud. A Private Cloud is suited for sensitive data, where the customer is dependent on a certain degree of security.
According to the different types of services offered, cloud computing can be considered to consist of three layers (services models): IaaS (infrastructure as a service), PaaS (platform as a service), SaaS (software as a service). Main cloud computing solutions: web applications, data hosting, virtualization, database clusters and terminal services. The advantage of cloud com-puting is the ability to virtualize and share resources among different applications with the objective for better server utilization and without a clustering solution, a service may fail at the moment the server crashes.
This work put forwards an optimal BCI (Brain Computer Interface) speller design based on Steady State Visual Evoked Potentials (SSVEP) and Artificial Neural Network (ANN) in order to help the people with severe motor impairments. This work is carried out to enhance the accuracy and communication rate of BCI system. To optimize the BCI system, the work has been divided into two steps: First, designing of an encoding technique to choose characters from the speller interface and the second is the development and implementation of feature extraction algorithm to acquire optimal features, which is used to train the BCI system for classification using neural network. Optimization of speller interface is focused on representation of character matrix and its designing parameters. Then again, a lot of deliberations made in order to optimize selection of features and user’s time window. Optimized system works nearly the same with the new user and gives character per minute (CPM) of 13 ± 2 with an average accuracy of 94.5% by choosing first two harmonics of power spectral density as the feature vectors and using the 2 second time window for each selection. Optimized BCI performs better with experienced users with an average accuracy of 95.1%. Such a good accuracy has not been reported before in account of fair enough CPM.