Biblio

Deep neural networks (DNNs) provide good performance for image recognition, speech recognition, and pattern recognition. However, a poisoning attack is a serious threat to DNN's security. The poisoning attack is a method to reduce the accuracy of DNN by adding malicious training data during DNN training process. In some situations such as a military, it may be necessary to drop only a chosen class of accuracy in the model. For example, if an attacker does not allow only nuclear facilities to be selectively recognized, it may be necessary to intentionally prevent UAV from correctly recognizing nuclear-related facilities. In this paper, we propose a selective poisoning attack that reduces the accuracy of only chosen class in the model. The proposed method reduces the accuracy of a chosen class in the model by training malicious training data corresponding to a chosen class, while maintaining the accuracy of the remaining classes. For experiment, we used tensorflow as a machine learning library and MNIST and CIFAR10 as datasets. Experimental results show that the proposed method can reduce the accuracy of the chosen class to 43.2% and 55.3% in MNIST and CIFAR10, while maintaining the accuracy of the remaining classes.

Deep neural networks (DNNs) exhibit excellent performance in machine learning tasks such as image recognition, pattern recognition, speech recognition, and intrusion detection. However, the usage of adversarial examples, which are intentionally corrupted by noise, can lead to misclassification. As adversarial examples are serious threats to DNNs, both adversarial attacks and methods of defending against adversarial examples have been continuously studied. Zero-day adversarial examples are created with new test data and are unknown to the classifier; hence, they represent a more significant threat to DNNs. To the best of our knowledge, there are no analytical studies in the literature of zero-day adversarial examples with a focus on attack and defense methods through experiments using several scenarios. Therefore, in this study, zero-day adversarial examples are practically analyzed with an emphasis on attack and defense methods through experiments using various scenarios composed of a fixed target model and an adaptive target model. The Carlini method was used for a state-of-the-art attack, while an adversarial training method was used as a typical defense method. We used the MNIST dataset and analyzed success rates of zero-day adversarial examples, average distortions, and recognition of original samples through several scenarios of fixed and adaptive target models. Experimental results demonstrate that changing the parameters of the target model in real time leads to resistance to adversarial examples in both the fixed and adaptive target models.

Recommendation systems become popular in our daily life. It is well known that the more the release of users' personal data, the better the quality of recommendation. However, such services raise serious privacy concerns for users. In this paper, focusing on matrix factorization-based recommendation systems, we propose the first privacy-preserving matrix factorization using fully homomorphic encryption. On inputs of encrypted users' ratings, our protocol performs matrix factorization over the encrypted data and returns encrypted outputs so that the recommendation system knows nothing on rating values and resulting user/item profiles. It provides a way to obfuscate the number and list of items a user rated without harming the accuracy of recommendation, and additionally protects recommender's tuning parameters for business benefit and allows the recommender to optimize the parameters for quality of service. To overcome performance degradation caused by the use of fully homomorphic encryption, we introduce a novel data structure to perform computations over encrypted vectors, which are essential operations for matrix factorization, through secure 2-party computation in part. With the data structure, the proposed protocol requires dozens of times less computation cost over those of previous works. Our experiments on a personal computer with 3.4 GHz 6-cores 64 GB RAM show that the proposed protocol runs in 1.5 minutes per iteration. It is more efficient than Nikolaenko et al.'s work proposed in CCS 2013, in which it took about 170 minutes on two servers with 1.9 GHz 16-cores 128 GB RAM.