Earticle

High-performance conjugate gradient (HPCG) is the latest benchmark adopted by the TOP500 organization, and thus how to optimize the HPCG source code for different heterogeneous computing platforms to achieve a higher floating-point computation rate has already become a new hot issue in HPC field. In the paper, we used the CPU + MIC heterogeneous computing platforms, and successfully ported HPCG to the platform. Through the analysis of HPCG source code and optimization for CPU + MIC platforms, practical significance and the value of further research is put forward. Results of performing the benchmark indicate that the design of optimization methods is reasonable and has facilitated the speedup of HPCG benchmark.

Deep learning is a technique of machine learning in artificial intelligence area. Deep learning is a refined "machine learning" algorithm that surpasses a considerable lot of its forerunners in its capacity to perceive syllables and pictures. As of now Deep learning is a greatly dynamic examination territory in machine learning and example acknowledgment society. It has increased colossal triumphs in an expansive zone of utilizations, for example, speech recognition, computer vision and natural language processing and numerous industry items. Neural networks are used to implement the machine learning or to design intelligent machines. In this paper thorough survey to all machine learning paradigms and application areas of deep machine learning and different types of neural networks with applications are discussed.

The challenge of modern machining industries is mainly to achieve high quality, in terms of dimensional accuracy, surface finish and high production rate. In the present work, an attempt has been made to explore the effect of EDM process parameters on various Surface Roughness characteristics Ra, Rq and Rz. The experiments have been done on medium carbon steel EN8 which has high applications in tool, oil and gas industries. For the experimentation, Pulse on time (PON), Pulse off time (POFF), Wire tension (WT) and Wire feed rate (WF) were considered as EDM inputs. Similarly, Surface roughness parameters Ra, Rq and Rz were taken as output characteristics. Taguchi’s standard L27 (3 level * 3 parameters) Orthogonal Array was used for conducting the experiments and optimization of multi responses was done by Grey analysis. From the Grey analysis, optimal combination of machining parameters for Surface Roughness characteristics was found at Pulse-on time (PON):131 μs (level 3), Pulse-off time (POFF):58 μs (level 2), Wire tension (WT):2 Kg-f (level 1) and Wire feed rate (WF):4 m/min (level 1). Analysis of variance was used to find the influence of process parameters on responses. From ANOVA of Grey Relational Grade, it is found that wire feed has high influence (F= 59.18) and pulse off time has very low influence (F= 2.50) in effecting the Surface Roughness parameters taken simultaneously. Regression models for the responses were prepared by using MINITAB-16 software. From the residual analysis, it is clear that the models prepared for the responses were more accurate, adequate and they can be used for prediction of surface roughness characteristics. Finally, the experimental and predicted values were compared and they found very close to each other.

E-Business has gained much attention as a new opportunity to provide global competitiveness for companies in developing countries. This study is primarily focused on critical success factors of e-Business systems in developing countries, especially in Kazakhstan. This study aims to provide a framework to unfold the barriers and challenges so that companies in Kazakhstan take benefit of making а right decision for e-Business implementation. To achieve the goal of the research, the survey was conducted to assess and understand the lаndsсарe of e-Business in Kazakhstan аnd identify the key suссess fасtors which can be utilized in improving the design and implementation of e-Business systems. Kazakhstan, as a leader of Central Asia, has a great potential to impact expansion of e-Business systems in this region. In order to prepare a well designed e-Business systems which reflect success factors identified in this study, a framework to develop a landscape design for e-Business systems might be necessary. This study revealed critical factors affecting successful implementation of e-Business in Kazakhstan, which has shown the reality of success factors in developing countries. It is expected that success factors identified in this study can be utilized in designing and implementing e-Business systems in developing countries. In addition, it might be desirable to stay aware of opportunities and challenges, and see it as an initiative to improve businesses rather than the solution, especially in developing countries.

Artificial neural network (ANN) have wide applicability in various applications in the life, one of them is apply to recognize characters and numbers, and we know that the Automatic License Plate Recognition (ALPR) is very important in the Intelligent Transportation System (ITS) and it is beginning in research and application in Vietnam. Usually, an ALPR system consists of three parts: 1) license plate location, 2) character segmentation, 3) characters & numbers recognition. In this paper, we proposed an improved method for the characters & numbers recognition part. And then, we apply to recognize characters & numbers of Vietnam License Plates (LP), which combined neural network and image processing technologies. In the training work, we used two networks and back-propagation (BP) algorithm for characters & numbers training with noises, separately, which the computing time and accuracy will be improved. In the using network work, we used the image processing technology for pre-processing to obtain high quality of characters & numbers before put in the trained network to improve accuracy of the system. We tested on 600 Vietnam LP images, which obtained from the actual systems, these images are very different background such as illumination, license angles, size and type, colors, light conditions in Vietnam environment. Our approach is more effective than some of the existing methods and satisfied for all types and color of Vietnam license plates and Vietnam environment.

The crystallization behavior and microstructural change in glass-ceramics with a cordierite composition were analyzed as a function of concentration and etching time of the HF solution in order to enhance the degree of crystallinity of the glass. Compared to a specimen without etching, the glass frit specimens etched with HF solution showed increased crystalline peak intensities of the μ-cordierite phase in the XRD pattern. The crystal size in the glass matrix also increased with an increasing concentration of the HF solution for the etched glass frit. The crystal peak intensity of the α-cordierite phase in the glass-ceramics made with glass frit etched with 0.5-1.5 wt% HF solution had a slightly higher value than that of pure glass frit regardless of the etching time. The intensity of the specimen with etched glass frit with 3.0 wt% HF solution, however, was lower than that of the specimen of the pure glass frit. Obviously, for the glass-ceramics etched with HF solution of various conditions, the α-cordierite phase is more favorable for crystallization than the α-cordierite phase. This study revealed that the microstructure and ratio of the crystallinity of the α-cordierite phase over the α-cordierite phase in the obtained glass-ceramics could be controlled by changing the HF-etching conditions followed by a heat treatment process.

Aimed at the defects of the traditional Canny operator, this paper puts forward an improved algorithm in edge detection. First, this paper gives a new method on selfadaptive image block based on threshold value. Next, by proposing a new hybrid filterbank of self-adaptive median and morphology, we adopt this hybrid filter-bank to smooth the noise image. Then, we add the information of gradient in two bevel directions, so that the information of gradient is more complete. Last, by using the threshold value to process the image of gradient which is after non-maxima suppression, we obtain the image edge. For the noise image, this improved algorithm not only can filter out noise well, but also the image edge is continuous, smooth, clear. The experimental results show that the improved algorithm has a good effect in edge detection, strong capability of noise immunity. The objective evaluation and visual effect are good, too.

The mobile game is experiencing a rapid development and one of the favorite mobile applications. This article focuses on Chinese mobile game market and conducts an empirical research on the factors of the user adoption of mobile game. The research results show that perceived usefulness, perceived entertainment, economic cost and subject norm significantly affect the attitude of use, which results in behavior intention to adopt the mobile game, while the perceived ease of use, similarity and brand trust, which proposed to affect the user adoption in related literatures insignificantly affect use adoption of mobile game .

In this paper, search simulation system is established to collect user clicking behavior. Empirical method is employed to study characteristics of results page which affect the user's clicks. Rankings, product brand information, retailer information, whether including product images, keywords length and other characteristics of the user click behavior are analyzed. The empirical results show that ranking is not the only factors that determine the result of click; well-known branded keyword or well-known retailer specific can improve click-through rate. Non well-known companies and non-famous retailer should get the top-ranking position to improve their ad clicks. User usually choose moderate length keyword to search. The article also firstly study the impact on commodity picture click-through rate (CTR). These conclusions have important guiding significance for advertisers to better carry out keyword ad activities.

How to manage the component library is important to the component reuse. The description mechanism of components is the basic on the component retrieval, which make users better understand the component. In general, we deal with three issues in semantic search, namely, usability, scalability and retrieval performance. In this paper, we propose a novel keyword-based semantic retrieval approach. Furthermore, the tradition component library exist heterogeneity because of the different company have their own standard. So the proposed methodology in the paper is fulfilled with the component management framework in the basic of the ontological feature model based on ontology and multiple agent system .The feature model is distinguish the tradition feature model, which describe the component form three different dimensions. Also the semantic retrieval model of component and the disturbed calling of component based on the multiple agents are researched which will help to improve the component reuse. The result indicates the effectiveness of the proposed framework.