人工智能原理姓名:成军学好:510061813论文题目:决策树算法在商标分类中的应用中文摘要:决策树一般都是自上而下的来生成的
每个决策或事件(即自然状态)都可能引出两个或多个事件,导致不同的结果,把这种决策分支画成图形很像一棵树的枝干
本文将使用决策树算法对给定的商标进行分类
其中有三大类商标数据,每大类使用五分之三的数据进行训练,使用五分之二的数据进行测试
我们应用 Java 和 MySQl 数据库进行测试
5 算法构造决策树
最终对数据进行准确率计算
关键词: 决策树 分类 商标 测试Title: Decision tree algorithm of the application for trademark classification Abstract: The decision tree is commonly top-down to generate
Each decision or events (namely natural state) are likely to elicit two or more events, lead to different results, put this decision branch loser graphics is like a tree branches
This article will use the decision tree algorithm for given the trademark of classification
There are three kinds of trademark data, each of the groups using three fifths of the data to train, use two fifths of the experimenta