TY - EJOU AU - Dong, Danping AU - Wu, Yue AU - Xiong, Lizhi AU - Xia, Zhihua TI - A Privacy Preserving Deep Linear Regression Scheme Based on Homomorphic Encryption T2 - Journal on Big Data PY - 2019 VL - 1 IS - 3 SN - 2579-0056 AB - This paper proposes a strategy for machine learning in the ciphertext domain. The data to be trained in the linear regression equation is encrypted by SHE homomorphic encryption, and then trained in the ciphertext domain. At the same time, it is guaranteed that the error of the training results between the ciphertext domain and the plaintext domain is in a controllable range. After the training, the ciphertext can be decrypted and restored to the original plaintext training data. KW - Linear regression KW - somewhat homomorphic encryption KW - machine learning DO - 10.32604/jbd.2019.08706