基于实例的机器学习方法研究
首发时间:2009-09-11
摘要:本文首先介绍了机器学习系统的基本结构,主要分析研究了基于实例的机器学习原理和特点,探讨了距离函数、有效寻找最近邻等机器学习方法,指出了这些方法存在的问题,通过减少样本集数量、修剪干扰样本集、属性加权和推广距离函数等真正实现了机器学习方法,并且得到了理想的结果。
For information in English, please click here
Research on Machine Learning Method Instance-based
Abstract:This article first introduces the basic structure of the machine learning system and mainly study principles and characteristics of machine learning based on instance-based, the distance function has been explored and looking for the nearest neighbor effectively has been analyzed deeply. The existence of problems in them has been pointed out. On the basis, real machine learning method has been achieved through reducing the number of sample collection, pruning interference sample, property weighted and distance function promoted, and obtains the desired results.
Keywords: machine learning instance distance function the nearest neighbor sample collection
基金:
论文图表:
引用
No.3512149063112526****
同行评议
共计0人参与
勘误表
基于实例的机器学习方法研究
评论
全部评论0/1000