摘要
逻辑回归模型作为分类算法已经被广泛应用到许多领域。近年来随着信息领域的高速发展,海量规模数据成为发展的主要趋势。传统构建逻辑回归模型的算法在大规模数据下不能有效地构建逻辑回归模型。针对海量数据,本文提出了高效的分布式逻辑回归模型构建算法。该算法是基于云计算平台,能够快速、高效地完成逻辑回归分类模型的构建。实验表明,本文提出的算法具有很好的加速比以及可扩展性。
Logistic regression model is applied into many areas as a classification algorithm.Recently,with the highly development of information field,high scale data is becoming the main development trend.Traditional building algorithms for logistic regression models could not build logistic regression models for huge scale data effectively.In this paper,focusing on huge scale data,we propose a high efficient distributed building algorithm for logistic regression models,and it is based on cloud computing platform,and it could build logistic regression classification models fast and efficiently.The experimental results show that the proposed algorithm in this paper has good speed-up and scalability.
出处
《科技通报》
北大核心
2013年第6期137-139,共3页
Bulletin of Science and Technology
基金
广东省高新技术产业化攻关项目(00595750177068012)
关键词
逻辑回归
云计算
分类
可扩展性
logistic regression
cloud computing
classification
scalability
