Inf. Sci.最新文献_第5页

Stateful Detection of Adversarial Reprogramming 对抗性重编程的状态检测

Inf. Sci. Pub Date : 2022-11-05 DOI: 10.48550/arXiv.2211.02885

Yang Zheng, Xiaoyi Feng, Zhaoqiang Xia, Xiaoyue Jiang, Maura Pintor, Ambra Demontis, B. Biggio, F. Roli

{"title":"Stateful Detection of Adversarial Reprogramming","authors":"Yang Zheng, Xiaoyi Feng, Zhaoqiang Xia, Xiaoyue Jiang, Maura Pintor, Ambra Demontis, B. Biggio, F. Roli","doi":"10.48550/arXiv.2211.02885","DOIUrl":"https://doi.org/10.48550/arXiv.2211.02885","url":null,"abstract":"Adversarial reprogramming allows stealing computational resources by repurposing machine learning models to perform a different task chosen by the attacker. For example, a model trained to recognize images of animals can be reprogrammed to recognize medical images by embedding an adversarial program in the images provided as inputs. This attack can be perpetrated even if the target model is a black box, supposed that the machine-learning model is provided as a service and the attacker can query the model and collect its outputs. So far, no defense has been demonstrated effective in this scenario. We show for the first time that this attack is detectable using stateful defenses, which store the queries made to the classifier and detect the abnormal cases in which they are similar. Once a malicious query is detected, the account of the user who made it can be blocked. Thus, the attacker must create many accounts to perpetrate the attack. To decrease this number, the attacker could create the adversarial program against a surrogate classifier and then fine-tune it by making few queries to the target model. In this scenario, the effectiveness of the stateful defense is reduced, but we show that it is still effective.","PeriodicalId":13641,"journal":{"name":"Inf. Sci.","volume":"21 1","pages":"119093"},"PeriodicalIF":0.0,"publicationDate":"2022-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78621414","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 2

Feature selection based on double-hierarchical and multiplication-optimal fusion measurement in fuzzy neighborhood rough sets 基于模糊邻域粗糙集双层次乘优融合测量的特征选择

Inf. Sci. Pub Date : 2022-11-01 DOI: 10.1016/j.ins.2022.10.133

Hongyuan Gou, Xianyong Zhang

引用次数: 3

Reward Shaping Using Convolutional Neural Network 基于卷积神经网络的奖励塑造

Inf. Sci. Pub Date : 2022-10-30 DOI: 10.48550/arXiv.2210.16956

Hani Sami, H. Otrok, J. Bentahar, A. Mourad, E. Damiani

{"title":"Reward Shaping Using Convolutional Neural Network","authors":"Hani Sami, H. Otrok, J. Bentahar, A. Mourad, E. Damiani","doi":"10.48550/arXiv.2210.16956","DOIUrl":"https://doi.org/10.48550/arXiv.2210.16956","url":null,"abstract":"In this paper, we propose Value Iteration Network for Reward Shaping (VIN-RS), a potential-based reward shaping mechanism using Convolutional Neural Network (CNN). The proposed VIN-RS embeds a CNN trained on computed labels using the message passing mechanism of the Hidden Markov Model. The CNN processes images or graphs of the environment to predict the shaping values. Recent work on reward shaping still has limitations towards training on a representation of the Markov Decision Process (MDP) and building an estimate of the transition matrix. The advantage of VIN-RS is to construct an effective potential function from an estimated MDP while automatically inferring the environment transition matrix. The proposed VIN-RS estimates the transition matrix through a self-learned convolution filter while extracting environment details from the input frames or sampled graphs. Due to (1) the previous success of using message passing for reward shaping; and (2) the CNN planning behavior, we use these messages to train the CNN of VIN-RS. Experiments are performed on tabular games, Atari 2600 and MuJoCo, for discrete and continuous action space. Our results illustrate promising improvements in the learning speed and maximum cumulative reward compared to the state-of-the-art.","PeriodicalId":13641,"journal":{"name":"Inf. Sci.","volume":"278 1","pages":"119481"},"PeriodicalIF":0.0,"publicationDate":"2022-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76819546","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Pruning Techniques in LinCbO for Computation of the Duquenne-Guigues Basis 用于Duquenne-Guigues基计算的LinCbO剪枝技术

Inf. Sci. Pub Date : 2022-10-01 DOI: 10.1007/978-3-030-77867-5_6

Radek Janostik, J. Konečný, Petr Krajča

引用次数: 3

Neighborhood-based differential evolution algorithm with direction induced strategy for the large-scale combined heat and power economic dispatch problem 大型热电联产经济调度问题的定向诱导邻域差分进化算法

Inf. Sci. Pub Date : 2022-09-01 DOI: 10.1016/j.ins.2022.09.025

Di Liu, Zhongbo Hu, Qinghua Su

引用次数: 10

Causality detection with matrix-based transfer entropy 基于矩阵传递熵的因果关系检测

Inf. Sci. Pub Date : 2022-09-01 DOI: 10.1016/j.ins.2022.09.037

Wanqi Zhou, Shujian Yu, Badong Chen