摘要
针对医学图像数据过于复杂且分布存储的特点,提出并实现了一种基于SLIQ的分布式医学图像分类框架。该框架包括:表示层、处理层和挖掘层。其中,分布式协调器(DTC)是处理层的核心,通过分析以往算法的优缺点,建立一种分布式数据挖掘的计算框架,并给出相应的求解算法。挖掘层中的分类算法采用适合处理海量数据的SLIQ决策树方法。实验结果表明该分类系统是有效和可行的。
Aiming at masses of medical image data is stored in the excessively complicated and distributed environment. This paper proposed and realized the architecture of distributed medial image classification based on SLIQ. The architecture includes:express layer, disposal layer and mining layer. In them, the distributed coordinator is the core in the disposal layer. The distributed Medical Image Classification architecture of data mining was established by analyzing mining theory merit & shortcoming of former algorithms. The classification algorithm was realized by the decision tree algorithm SLIQ in mining layer. The experiments show that the method is effective and feasible.
出处
《微计算机信息》
北大核心
2008年第15期309-311,共3页
Control & Automation
基金
国家自然科学基金(60572112)
关键词
分布式数据挖掘
医学图像
分类算法
Distribute Data Mining
Medical Image
classification algorithm