Privacy and Security Issues in Deep Learning: A Survey

Liu, Ximeng; Xie, Lehui; Wang, Yaopeng; Zou, Jian; Xiong, Jinbo; Ying, Zuobin; Vasilakos, Athanasios V.

doi:10.1109/access.2020.3045078

Cited by 189 publications

(91 citation statements)

References 123 publications

(161 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Four mainstream technologies for privacy protection from DL are currently available, namely differential privacy, homomorphic encryption, reliable multi-part computing and a trustworthy running environment [149]. Differential privacy is intended to prevent an opponent from finding out if a specific case had trained the target model.…”

Section: And Explainability: a Machine Learning Zoo Mini-tour And Explainable Ai A Review Of Machine Learning Interpretability Methods Pamentioning

confidence: 99%

On Deep Research Problems in Deep Learning

Ulhaq¹

2021

Preprint

View full text Add to dashboard Cite

The subject of deep learning has emerged in the last decade as one of the most promising approaches to machine learning. Today, certainly, much of the recent progress in artificial intelligence is due to it, but research challenges are still unresolved and remain open to the research community. This paper attempts to offer a comprehensive review of deep learning progress in active research frontiers. On the one side, by presenting a brief overview of deep learning success, we inspire researchers to work in deep learning. On the other hand, we examine a range of technical issues, and open research issues that we believe are relevant topics for exploratory research. As deep learning applies to various fields, we restrict this paper’s scope to visual recognition tasks to analyze these problems with a specific lens. However, these problems will be broadly applicable to other fields. It will make it easier for new researchers to recognize outstanding research problems in the deep learning domain.

show abstract

Section: And Explainability: a Machine Learning Zoo Mini-tour And Explainable Ai A Review Of Machine Learning Interpretability Methods Pamentioning

confidence: 99%

On Deep Research Problems in Deep Learning

Ulhaq¹

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Machine learning algorithms have been used to avoid or detect attacks and security problems, including cloud vulnerabilities, in a variety of ways [56]. The use of machine learning and its applications in cloud computing and related environments has been discussed in some of the most recent related works [57][58][59][60][61].…”

Section: Cloud Computing Replication Strategiesmentioning

confidence: 99%

Quantitative Analysis and Performance Evaluation of Target-Oriented Replication Strategies in Cloud Computing

Waseem¹,

Din

Alshamrani

et al. 2021

Electronics

View full text Add to dashboard Cite

Data replications effectively replicate the same data to various multiple locations to accomplish the objective of zero loss of information in case of failures without any downtown. Dynamic data replication strategies (providing run time location of replicas) in clouds should optimize the key performance indicator parameters, like response time, reliability, availability, scalability, cost, availability, performance, etc. To fulfill these objectives, various state-of-the-art dynamic data replication strategies has been proposed, based on several criteria and reported in the literature along with advantages and disadvantages. This paper provides a quantitative analysis and performance evaluation of target-oriented replication strategies based on target objectives. In this paper, we will try to find out which target objective is most addressed, which are average addressed, and which are least addressed in target-oriented replication strategies. The paper also includes a detailed discussion about the challenges, issues, and future research directions. This comprehensive analysis and performance evaluation based-work will open a new door for researchers in the field of cloud computing and will be helpful for further development of cloud-based dynamic data replication strategies to develop a technique that will address all attributes (Target Objectives) effectively in one replication strategy.

show abstract

“…The integration of DL in reduced the curse of dimensionality and improved the ability to solve high problems [22]. The security of DL models and data privacy protection has be [23,24] and remains a concern to service providers, especially with the massi ity of objects. These security threats can lead to inaccurate models for tra agents and adversely affect performance especially for time-bound applicatio mises to data through network attacks [23,24] lead to the availability of se mation to hackers and the occurrence of this phenomenon will compromise tiality of the system.…”

Section: Introductionmentioning

confidence: 99%

“…The integration of DL in RL has also reduced the curse of dimensionality and improved the ability to solve high-dimensional problems [22]. The security of DL models and data privacy protection has been studied in [23,24] and remains a concern to service providers, especially with the massive connectivity of objects. These security threats can lead to inaccurate models for training the RL agents and adversely affect performance especially for time-bound applications.…”

Section: Introductionmentioning

confidence: 99%

“…These security threats can lead to inaccurate models for training the RL agents and adversely affect performance especially for time-bound applications. Compromises to data through network attacks [23,24] lead to the availability of sensitive information to hackers and the occurrence of this phenomenon will compromise the confidentiality of the system.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Multi-Agent Reinforcement Learning Framework in SDN-IoT for Transient Load Detection and Prevention

Dake¹,

Gadze²,

Klogo³

et al. 2021

Technologies

View full text Add to dashboard Cite

The fast emergence of IoT devices and its accompanying big and complex data has necessitated a shift from the traditional networking architecture to software-defined networks (SDNs) in recent times. Routing optimization and DDoS protection in the network has become a necessity for mobile network operators in maintaining a good QoS and QoE for customers. Inspired by the recent advancement in Machine Learning and Deep Reinforcement Learning (DRL), we propose a novel MADDPG integrated Multiagent framework in SDN for efficient multipath routing optimization and malicious DDoS traffic detection and prevention in the network. The two MARL agents cooperate within the same environment to accomplish network optimization task within a shorter time. The state, action, and reward of the proposed framework were further modelled mathematically using the Markov Decision Process (MDP) and later integrated into the MADDPG algorithm. We compared the proposed MADDPG-based framework to DDPG for network metrics: delay, jitter, packet loss rate, bandwidth usage, and intrusion detection. The results show a significant improvement in network metrics with the two agents.

show abstract

Privacy and Security Issues in Deep Learning: A Survey

Cited by 189 publications

References 123 publications

On Deep Research Problems in Deep Learning

On Deep Research Problems in Deep Learning

Quantitative Analysis and Performance Evaluation of Target-Oriented Replication Strategies in Cloud Computing

Multi-Agent Reinforcement Learning Framework in SDN-IoT for Transient Load Detection and Prevention

Contact Info

Product

Resources

About