作者机构：
[Sui Wu] School of Music,Hunan City University
会议名称：
2014 2nd International Conference in Humanities,Social Sciences and Global Business Management(ISSGBM 2014)
会议时间：
2014-06-21
会议地点：
London,UK
摘要：
In recent ten years, on the Internet to solicit songs is very active, according to preliminary statistics, no less than 100 years, the large scale, the many of number, bonus high, has formed a kind of
摘要：
the differential equations of different classification, and image registration as the application object, the differential equations of different types of modeling finally, in different images in diff
摘要：
This paper combines geometric flow method with tessellation and make full use of their respective strengths to complete some surface design problems, such as surface blending, N sides fill holes and o
摘要：
In order to solve the problems of multi-layer network energy consumption, for the multi-layer network with an IP layer, TDM layer and optical layers 3-layer structure, ITO model is proposed in this pa
作者机构：
[Hean Liu] College of Science, Hunan City University
会议时间：
2017-10
摘要：
classroom teaching is the main position for teachers and students to carry out teaching activities. It is also the position that teachers and students influence each other and grow in parallel. We adv
作者机构：
[Hean LIU; Jianjun WU] Hunan City University
会议时间：
2017-08
摘要：
Aiming at the problems in service oriented real-time multimedia transmission, such as a long delay, frequent jitter and low reliability, this paper proposes LD/RPath algorithm LD/RPath estimates the d
摘要：
The theoretical research and practical application of internal control experienced from single scattered to comprehensive development process. There is no doubt that this is closely related to the cha
摘要：
To solve the problem like calculation of migration frequently existing in the cloud computing system, storage migration, backup computing, storage, backup and node failure handling high coupling between nodes operating characteristics, this paper presents analysis model based on phase space cloud scheduling algorithms. This article makes cloud computing system project to phase-space, defining and analyzing the generalized phase space temperature, generalized normalized entropy, generalized gravity, generalized normalized momentum, generalized thermodynamic parameters etc. Taking advantage of thermodynamic principles to establish a cloud computing system phase space analysis model of the basic theoretical framework, taking advantage of phase space analysis model scheduling algorithm for cloud computing cloud computing can be projected to the parameter phase space and momentum phase space. Experimental results show that: the proposed algorithm can effectively make cloud computing system achieve load balancing in phase space remain thermodynamically non-equilibrium systems.
摘要：
As the ART2 neural network clustering occurs normalization in the data inputting mode by vector and nonlinear transformation pretreatment process is easy to be filtered as a substrate for an important, but a minor component of the noise, while there are still phenomenon of the drifting mode in the learning process due to the correction of the value of weight, this paper proposes an improved method of ART2 neural network. The improved method stores the amplitude information in the learning process, and it is considering the shortest distance of being inputted into the center of the cluster, increasing a threshold limit value for determining outliers at the same time and eliminating the influence of outliers of the clustering results. Finally, the clustering of data samples experimental results show that: the improved ART2 network can handle negative data, the four quadrants of data can be effectively clustered, the performance is superior to the traditional ART2 network.
