Time filter

Source Type

Jin Y.,Shi Jiazhuang Post and Telecommunication Technical College | Geng K.,Shi Jiazhuang Post and Telecommunication Technical College | Wang Y.,Shi Jiazhuang Post and Telecommunication Technical College | Zhao B.,Shi Jiazhuang Post and Telecommunication Technical College
International Journal of Advancements in Computing Technology | Year: 2012

To improve the face recognition rate and reduce the complexity of the calculation, an efficient feature reduction algorithm based on modular principal component analysis (mPCA) and rough set. To cope with the variations of pose, lighting direction and facial expression, first divided the face images into smaller sub-images, and then used PCA method to deal with them. To guarantee that the selected first principal components as a feature vector, using rough sets to simplified the features of the processed sub- images by mPCA. At the same time, get rid of the extra, not related and redundant data, to reduce the complexity of the time and space. Then the results indicate high improvement in the classification performance compared to the conventional method. The innovation point of the paper is the perfect combination of the mPCA methods and rough sets. The experiment results show the validity of this method.


Jin Y.,Shi Jiazhuang Post and Telecommunication Technical College | Gu B.,Shi Jiazhuang Post and Telecommunication Technical College | Wang Y.,Shi Jiazhuang Post and Telecommunication Technical College | Wang Q.,Shi Jiazhuang Post and Telecommunication Technical College
Journal of Convergence Information Technology | Year: 2012

To solve the quantification problem in Text information mining, proposed a novel method of knowledge discover based on certainty factor model. Firstly, three databases of interest, industry and advertising are built based on magazine and its subscription information. Then input the uncertain information into certainty factor system. The output data show the CF value of interest, industry of Magazine subscriber and the CF value of advertising may published in any magazines. Also based on the certainty factor system, the value of the certainty factor with some comprehensive conditions can be calculated. The conclusion could be applied to direct mail database marketing to get a better result.


Jin Y.,Shi Jiazhuang Post and Telecommunication Technical College | Wang Y.,Shi Jiazhuang Post and Telecommunication Technical College | Geng K.,Shi Jiazhuang Post and Telecommunication Technical College | Zhao B.,Shi Jiazhuang Post and Telecommunication Technical College
Lecture Notes in Electrical Engineering | Year: 2013

The certainty factor is an inaccuracy inference model used by MYCIN system. It is a reasonable and effective inference model for many practical applications. This paper will focus on the analysis of text messages of magazines and build the audiences' interest, keywords of their careers. Based on the certainty factor, we can calculate the value of the certainty factor with some comprehensive conditions, and then learn the audiences' interest, the level of the certainty factor for their careers with the value in different conditions. This conclusion could be applied to direct mail database marketing to get a better result. © 2013 Springer-Verlag.


Yanfeng J.,Shi Jiazhuang Post and Telecommunication Technical College | Bingtong W.,Shi Jiazhuang Post and Telecommunication Technical College | Bin G.,Shi Jiazhuang Post and Telecommunication Technical College | Kai W.,Shi Jiazhuang Post and Telecommunication Technical College
Lecture Notes in Electrical Engineering | Year: 2014

A novel algorithm for face recognition based on principal components analysis (PCA) and rough set is presented in this paper. First the face images are divided into smaller subimages and the PCA approach is applied to each of these subimages.We expect the proposed method to be able to cope with these variations of the pose, lighting direction, and facial expression. However, PCA does not guarantee that selected first principal components, as a feature vector, will be adequate for classification. Rough sets assume that a processed data set contains patterns labeled by associated classes, with the discrete values of its elements (attributes, features). Thus after realizing the KLT, resulting projected patterns features must be discretized by some adequate procedure. The resulting discrete attribute valued data set (an information system) can be processed using rough set methods. Then the results indicate high improvement in the classification performance compared to the conventional method. © Springer-Verlag Berlin Heidelberg 2014.


Wang Y.,Shi Jiazhuang Post and Telecommunication Technical College | Jin Y.,Shi Jiazhuang Post and Telecommunication Technical College | Li Y.,Shi Jiazhuang Post and Telecommunication Technical College | Geng K.,Shi Jiazhuang Post and Telecommunication Technical College
Communications in Computer and Information Science | Year: 2013

Association rules are the key technology in data mining; it has a very broad applying foreground in many industries. An improved association rules algorithms based on Apriori was proposed in this paper. And it will be used in direct mail data mining. By analyzing the normative database of users' sets, we can get item set which satisfy the minimal support degree, and form the rule set. We can get more accurate DM data mining results than other methods by testing the post DM database. Experiments indicate the validity of the method. © Springer-Verlag Berlin Heidelberg 2013.


Zhao B.,Shi Jiazhuang Post and Telecommunication Technical College | Zhang H.,Shi Jiazhuang Post and Telecommunication Technical College | Jin Y.,Shi Jiazhuang Post and Telecommunication Technical College | Su Y.,Shi Jiazhuang Post and Telecommunication Technical College
Communications in Computer and Information Science | Year: 2013

Polygon triangulation cutting algorithm has been widely used because of its excellent characteristics. The method is introduced into route standardization in this paper. Firstly, triangulate the polygon composed by Geometric points with Displacement and time by using Polygon triangulation cutting algorithm. Use the formula to calculate division of the triangle area, and then compute the area of the whole polygon. We can get n different polygons by scanning different period, repeat the algorithm and we can get the area of n polygons. Through the weighted average method to get the best a group of polygon set point, which composed the best path. The results of the experiments prove the validity of the method. © Springer-Verlag Berlin Heidelberg 2013.

Loading Shi Jiazhuang Post and Telecommunication Technical College collaborators
Loading Shi Jiazhuang Post and Telecommunication Technical College collaborators