The Research Of The Factor Analysis And Implementation In Afc Data Analysis System

Statistical analysis software can be applied to analize existing business data. It can discover latent rules and relationships among the data and support effectively business analysis for various industries, and the effective helps could be given to decisionmakers. Normal statistic software performs all of the work such as data storage, data preparing, statistic calculating and result displaying on one computer pattern. This kind of architecture pattern can't meet actual application requirements in today of network application developing high.The necessities and relative techniques' advantages for designing and developing statistical analysis software established on 3-tier architecture pattern have been analyzed in this paper based on researching and analyzing current major statistic softwares. A data analysis system has been designed and developed which is suitable for native situation of China and the industry characteristics of automatic fare collection (in short form AFC). The statistical analysis software based on 3-tier architecture researched and realized which has the property of distributed processing technique, in this paper can set in facility all the items such as data extracting, preparing, statistic calculating and results' displaying on different computing resources in network so that network computing resources can be used in sufficiency and computing performances can be enhanced.The primary research work described in this paper is the followings:1) key techniques for data extracting from isomeric data sources on cross-platforms have been solved by adopting the concept model of meta-data through analyzing data extracting techniques;2) Semantic descriptions in Chinese and their applications of relative information in meat-data have been proposed in terms of the concept model of meta-data, so that the data structure information of data market can be displayed with canonical industry terms and facility to select data object that might be analyzed;3) Business database resources' consumption has been reduced by splitting some relationship operating and statistic calculating from the host database system;4) The storing momentarily strategy of statistic result with the time stamp is proposed so that the subsequent requests can be responsed within permission time range;5) The interface protocols between the layers have been defined in terms of XML format files so that applicabilities and modifies can be provided with;6) The algorithm of factor analysis had been studied and realized, bring forward improvement algorithm of the factor analysis based on singular value decomposition. The factors are searched by combining singular value decomposition and principal component analysis, and applying to the AFC terminal equipments' trouble affair analysis of rail transit.The achievements mentioned above of 1) to 5) are participated in and 6) is completed independently by author.The ultimate purpose of researches in this paper is to design and realize "Automatic fare collection (AFC) operation management data analysis system of rail transit", which is checked and accepted by Shanghai Science and Technology Committee on July 12th, 2007, and registered and awarded the certificate of science and technology achievement (Registered No. : 9312007Y1168) on August 7th, 2007, and the certificate of computer software copyright (Registered No. : 2007SR13214) by the National Copyright Bureau on August 30th, 2007.