中文摘要：

企业间进行合作时，经常涉及到对来自多个企业的数据信息进行分析。在某些情况下，由于这些数据涉及到各自的私有信息，如果不能保护这种私有信息，企业不愿意公开这些数据信息，但可能愿意提供这些相关数据的总体特征。基于这样的思路，针对企业数据的同构和异构两种不同分布方式，以及Paillier同态加密的两个主要性质，对合作企业的数据进行隐私计算，从而实现了企业间私有信息的合作线性回归分析。

英文摘要：

The cooperative statistical analysis on data sets of various enterprises is often done during enterprises＇ cooperation. In certain circumstances, since the data may involve private information, enterprises are unwilling to share their data in the absence of data protection, but they are willing to provide the whole characteristics of their data. Based on the thought, with Paillier＇ s homomorphic eneryption, enterprises＇ sensitive data with isomorphic and heterogeneous distributions are computed. Consequently, the collaborative linear regression analysis with enterprises＇ private information protection is solved.