The Data Mining Method And Design Based On The Multiple Regression Equation Fitting

Statistical analysis refers to the use of statistical methods and analysis ofobject-related knowledge from a combination of quantitative and qualitative researchactivities. The statistical analysis of the more popular is the regression analysis than inthe past simple statistics to be accurate. But has its own advantages and limitations ofstatistical analysis, the first is the complexity of real life data, relying on statisticalmethods to control and interpret these complex data is not comprehensive; followedby statistical analysis based on probability, since the probability will be errors, it isdifficult to determine a statistical conclusions are absolutely correct.In recent years, data mining has caused in the information industry, great attention,it is an emerging area of research, the use of advances in artificial intelligence andstatistical analysis, as well as the application database and machine learning, byanalyzing a large number of random data find implicit regular data, and potentiallyuseful information, useful information can be widely used in e-commerce, marketanalysis and decision support.But data mining is not to deny the traditional statistical analysis, it is an extensionand expansion of the statistical analysis.In manual telephone system, when all of the phone is busy, the next time youenter the phone had been prompted to a busy state an expected wait time informationrather than feedback. Separate regression analysis method to solve this problem is onefor each phone, and tap out the characteristic value of this phone call the number up tothat point in time, calculate the current into the telephone immediately after therelated calculations are expected to wait for time.Around the traditional statistical analysis, regression analysis method to improvethe method by fitting multiple regression equation, so that the handover accuracy hasimproved to some extent, and this method is applied to the popular data miningsystem, the statistical theory to better serve the theory of data mining.