数据分类中决策树算法的一些改进

An Improved Decision Tree Algorithm in Data Classification

分类是数据挖掘中重要的研究课题。决策树方法是一种常用的分类算法,所建立的树型结构模型很直观,易于理解。传统的分类方法在处理海量数据时会出现性能下降或精度降低的问题,经过改进的ID3算法,基于SPRINT,消除了内存的限制,运算速度快,具有可伸缩性,性能较好。

Classification is an important topic in data mining research. Decision tree is one of well-known classification algorithms with its tree model easy to understand. However, traditional decision tree classification algorithms have low performance or low precision when processing large data sets. Based on SPRINT, an improved decision tree algorithm （ID3） reduces memory footprint and improves performance with comparable sealability.