【摘要】 提出了一种针对流数据的增量式多维可扩展可视化挖掘方法(Incremental multi-di-mension scaling,IMDS),对数据表现的特征形状进行聚类,并且聚类结果会随着时间的推移用动态可视化的方式实时展现。仿真实验表明:本文算法相比传统的MDS(Multi-dimension scaling)算法和简易型SIMPLEX优化算法在流数据挖掘中可以明显地提高可视化挖掘效率和流挖掘效果。更多还原

【Abstract】 An Incremental Multi-Dimension Scaling(IMDS) visualization method is proposed to discover patterns in quantities of data.IMDS algorithm clusters data through the shape of each structure data instead of traditional cluster algorithm,which needs global information for precision category result.Moreover,IMDS algorithm can be implemented through animation because that it maps multi-dimension to low dimension.By experiments,it is proved that the proposed IMDS algorithm outperforms Multi-Dimension Scaling(MDS) and simplex algorithm in efficiency and effectiveness.更多还原