摘要
分类是机器学习领域最重要的一类问题,其中K近邻法和Logistic回归是两个重要的机器学习算法。本文主要研究了K近邻算法和Logistic回归模型在数据分类问题中的具体应用。针对K近邻算法,在考虑数据特征基础上,分别用欧氏距离和曼哈顿距离作为距离度量,同时,对于Logistic回归分类问题,提出了一种改进的随机梯度上升算法。通过选取了UCI机器学习数据集中的Horse Colic、Wine Quality两个数据集对算法进行验证,应用结果表明:K近邻算法中使用欧氏距离更适合Wine Quality,并且改进的随机梯度上升算法显著提高了Logistic学习机器的训练时间,说明了K近邻法和改进Logistic回归分类算法具有良好的分类效果。
Classification is the most important issue in the field of machine learning. K-Nearest Neighbor algorithm and Logistic regression are two important machine learning algorithms. This paper studies the application of the K-Nearest Neighbor algorithm and a Logistic regression model in data classification. On the basis of date feature,respectively taking Euclidean distance and the Manhattan distance as distance metrics in K-Nearest Neighbor method,meanwhile,an improved stochastic gradient ascent algorithm has been put forward in the classification issue of Logistic regression. Through the selection of Horse Colic and Wine Quality from UCI machine learning data sets,this algorithm is verified. Results show that Euclidean distance is more suitable for classification of Wine Quality in K-Nearest Neighbor algorithm,and improved stochastic gradient ascent algorithm significantly improves Logistic training time,further illustrates that K-Nearest Neighbor algorithm and Logistic regression has an excellent classification effects.
出处
《洛阳理工学院学报（自然科学版）》
2016年第3期83-86,93,共5页
Journal of Luoyang Institute of Science and Technology：Natural Science Edition
基金
国家自然科学基金项目(41272363)
