Establishing Automatic Prediction System in Data Mining Process

• Main Authors: Yeh Shao-Wei, 葉劭緯
• Other Authors: 謝邦昌
• Format: Others
• Language: zh-TW
• Published: 2005
• Online Access: http://ndltd.ncl.edu.tw/handle/61979617289328812284

碩士 === 輔仁大學 === 應用統計學研究所 === 93 === Following the development of information science and technology, how to already to fully utilize information which we had in business competition is very important, keeping key point has become enterprises' goals. Because the improvement of computer science and technology, make enterprises improve the convenience about accumulating customer's characteristics, in order to obtain the advantage on the market with keen competition, enterprises will collect any information and behavior of the customer through expanding the computer equipment constantly, expect to hold old customers, and find new customer, Through using neural network, algorithms and machine learning methodology to analyze the collected data. Traditional statistics analytical methodologies are also applied to analyze in large data set, and explore useful characteristic, relationship and hiding patterns in large data set through analysis, Finally, we can provide exploring information to enterprises as the reference of generating strategies. When we analyzed data, Historical data sets are very useful and powerful information sources to analysis and explore hiding patterns. This research's main goal is to develop an automatic prediction system by visual basic program language for time series data set such as stocks, electricity predicting in data mining procedure. To help analyzer explore hiding patterns or establish predicted model quickly.