Privacy data security has become an important bottleneck for the overall development of artificial intelligence and a key challenge that needs to be broken in the Internet era. The current research mainly considers differential privacy to effectively protect the private information in the data. However, as the noise increases, the precision of the training model will decrease. In order to solve above problem, an adaptive differential privacy (ADP) method is constructed and applied to deep neural networks.ADP adds noise adaptively in the training process according to the importance of features. We also build the differential privacy multi-objective optimization model (DPMOM). DPMOM adopts multi-objective optimization characteristics, takes accuracy and privacy protection as the optimization objectives. It optimizes the super parameters of deep neural networks and the noise of differential privacy. In addition, to better solve the ADP model, with the NSGA-II algorithm as the basic framework, a multi-objective optimization algorithm based on differential privacy protection (DPP-MOA) is designed. Simulation experiments show that compared with other machine learning methods and differentially private stochastic gradient descent, the accuracy of ADP is higher under the same amount of noise. Through comparison with NSGA-II, IBEA, PESA-II, and AGE-II, DPPMOA is proved that the solution set of this algorithm is better. K E Y W O R D S deep neural networks, differential privacy protection, multi-objective optimization algorithm, multi-objective optimization model
INTRODUCTIONPrivate information under the Internet has the characteristics of large amount of data, multiple categories, and complex hierarchical relationships. The existing privacy protection technologies based on anonymization and data encryption need to closely rely on background knowledge assumptions, 1 which can only ensure that privacy on a single data set is not leaked. The limitations are often difficult to meet the requirements for big data privacy protection under the Internet. Therefore, people began to pay attention to privacy protection technologies and their applications such as secure multi-party computing, homomorphic encryption, and differential privacy.K-anonymity 2 protects user privacy by anonymizing data. To prevent quasi-identifiers from forming correspondence with other sets containing key attributes and quasi-identifiers, the key attributes are deleted before the data table is published and the quasi-identifiers are anonymized, making each record indistinguishable from the other records. However, data are protected by deleting the identifier of the traditional data before the data table is released, and the data table is vulnerable to chain attacks, causing user privacy exposure. As an important research direction of cryptography, secure multi-party computation was first proposed by Yao 3 to solve the problem of a group of mutually distrustful participants each holding