The Approach of Adaptive Spectral Clustering Analyze on High Dimensional Data

Data mining is an important tool in knowledge discovery. It aims at discovering the valuable patterns hidden in large volume of data. In a distributed environment, the main problem we need to consider for data mining is how to transfer the minimal amount of data and provide maximum sharing of information with the continued expansion of business scale and constant update of services content. For high-dimensional scientific data, an adaptive spectral clustering method has been proposed. The experimental results on numerical simulation of scientific data have shown the improvement of proposed method.