Application Research Of Cluster Analysis Based On Key Indicators In School Project Fund Allocation

With the highly application of the computer technology and information technology, there are amount of data information produced and stored in the EXCEL form or database management system in the field of education. Although information technology and database technology solved the problems from paper data into electronic data Information, a large number of accumulated data information lead a low utilization of large amounts of data for us, so that it is difficult for us to extract valuable information in limited time. Instead of these useful or useless data, we only use them for archive and records. Therefore, needs of data analysis and mining work become increasingly important.Based on the present situation of education funds allocation, how each sum of every school a more scientific and reasonable allocation of funds, in the face of every year is education management departments and solve the problem. Therefore, this article will be based on the current situation and the school of basic data, using data mining in the condensed hierarchical clustering method, the demand for money based data analysis and clustering of data in table, according to an interactive system, artificial intervention of hierarchical clustering results, end up with a satisfactory capital allocation scheme.Based on key indicators for the application of clustering analysis in founds allocation can be divided into three main processes:1. Preprocessing to the found allocation based data tables, to get the data source befitting for data mining.2. According to the preprocessing data, choose a suitable distance calculation method to calculate the similarity between each two schools, and clustering.3. Develop an interactive system with hierarchical clustering algorithm, after clustering analysis to any area of schools, artificial intervention the results of clustering, and get a satisfying founds allocation.Experimental results show that the capital allocation of interactive system, the capital allocation based data analysis results and in previous years, the actual capital allocation result accuracy reached 90%. By condensing hierarchical clustering method is applied to capital allocation basis data for research, not only improving the efficiency of the data, also reduces the human factors of interference degree of the allocation of funds, to a great extent, help policy makers to solve the problem of capital allocation and confusion, provides a scientific and effective decision-making basis for policymakers.