基于余弦相似度和实例加权改进的贝叶斯算法①.PDF

计 算 机 系 统 应 用 2016 年 第 25 卷 第 8 期 ① 基于余弦相似度和实例加权改进的贝叶斯算法 王行甫, 付欢欢, 王 琳 ( 中国科学技术大学 计算机学院, 合肥 230027) 摘 要: 面对大量样本特征时很多分类器无法取得较好的分类效果, 样本数有限导致贝叶斯算法无法获得精确的 联合概率分布估计, 在样本局部构建高质量分类器需要有效的样本相似性度量指标. 针对以上问题, 提出了一种基 于余弦相似度进行实例加权改进的朴素贝叶斯分类算法. 算法考虑特征对分类的决策权重不同, 使用余弦相似度度 量样本的相似性, 选出最优训练样本子集, 用相似度值作为训练样本的权值来训练修正后的贝叶斯模型进行分类. 基于 UCI 数据集的对比实验结果表明, 提出的改进算法易于实现且具有更高的平均分类准确率. 关键词: 实例加权; 朴素贝叶斯; 余弦相似度; 逆文本频率; 文本分类 Improved Naïve Bayes Algorithm Based on Weighted Instance with Cosine Similarity WANG Xing-Fu, FU Huan-Huan, WANG Lin (School of Computer Science and Technology, University of Science and Technology of China, Hefei 230027, China) Abstract: Many classifiers cannot get good results facing numerous sample features, bayes algorithms get poor estimate of joint probability distribution with limited samples, effective similarity measure is needed to build a local classifier. Considering these problems, an improved multinomial Naïve Bayes algorithm based on weighted instances (IWIMNB) with cosine similarity is proposed. Taking different attributes contributing differently to the classification decision weight into account this algorithm uses cosine similarity as a metric of the similarity between training and validation instances. Selected training instances weighted by cosine similarity are used to train modified Naïve Bayes model. Results of final experiments show that the average classification accuracy of the proposed IWIMNB algorithm gets significant improvement. Key words: weighted instance; Naïve Bayes; cosine similarity; inverse document frequency; text categorization 文本分类的主要任务是预先提供一定量已标记类 为了提高贝叶斯算法的分类效果, 文献[2,3,4]中 别的训练文本, 根据测试文本的内容判定其真实类别, 主要提出了三类改进方法: 1) 结构拓展法, 用有向边