Research And Application On Neural Network In Data Warehouse

With the acceleration of national is higher and education expansion in scale, thedegree of information management in Colleges and universities is more and moreadvanced, the talents from the university are playing important roles to the socialdevelopment. Therefore, colleges and universities play an important role in trainingqualified-national builders. Numbers of data come out when the rapid developmentand wide application of information technology, the scale of information managementin College and university is more and more bigger. Scientific and rational applicationof data mining is used in the college’s data analysis, and can find the relationship andspecific rules, it will play a great role to the construction and talent cultivation.First of all, the data warehouse and data mining technology are discussed in thispaper, we will introduce the implementation process of the data warehouse and datamining technology and methods in data mining.This topic is the concrete applicationof data mining technology in College Students’ scores. In order to cooperate withresearch, we combine the existing resources and rules with an university in Shenyang,and design a system with the J2EE technology. The students’ scores are the object ofthe office automation system. With the collection to the score of students, we canreflect some influential factors of students’ test scores in a certain extent.Secondly, this article takes the students of the institute of information as the mainobject in this system, and uses the BP neural network in data mining to analysisteaching resources, and predict some professional courses’ score.There are some jobsto do in this period also, including gather the factors of influence to the test results,select and deal with the original data, construct the structure of neural network. Thearticle will optimize the network too, a new BP neural network with double hide-layerand momentum item to predict results, and then analysis and comparison theprediction results. Finally, through the analysis to the forecast results, it must consider some mattersthat are needed to attention in designing and implementing class in the process of thefuture curriculum. In order to provide references to further improving teaching level.