Towards developing a data security aware federated training framework in multi-modal contested environments

Ovi, Pretom Roy; Dey, Emon Kumar; Roy, Nirmalya; Gangopadhyay, Aryya; Erbacher, Robert F.

doi:10.1117/12.2618904

Cited by 11 publications

(11 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Sensitive fields such as healthcare and business need to keep their data in a protected manner. The cloud-based machine learning (Jamal and Wimmer, 2023) and federated learning approach (Ovi et al, 2022) has created a secure environment for AI.…”

Section: Ethical Implications Of Using Chatgpt and Ai Technologiesmentioning

confidence: 99%

ChatGPT as an important tool in organizational management: A review of the literature

Ayinde,

Wibowo,

Ravuri

et al. 2023

Business Information Review

View full text Add to dashboard Cite

ChatGPT is an emerging technology that revolutionizes organizational practices, fundamentally altering how individuals in organizations search for, generate, and utilize information within the workplace. The effective functioning of many organizations heavily relies on data and information. ChatGPT has streamlined the process of working with data and information, making it more accessible for organizations and the individuals involved. However, the adoption of ChatGPT also presents challenges across various domains, including social, economic, and legal considerations. This study conducts a comprehensive literature review to explore multiple perspectives on integrating ChatGPT into organizational management. It examines various aspects, such as the development of ChatGPT, its practical uses, ethical implications, governance mechanisms, regulations and policies. This study aims to guide managers and stakeholders in effectively incorporating ChatGPT into their organizational processes. This study provides a detailed examination of ChatGPT's impact on organizational management, offering valuable insights to practitioners and scholars alike, aiming to navigate the complexities and harness the benefits of this transformative technology. By understanding the implications and leveraging the potential of ChatGPT, organizations can enhance their operations and decision-making processes.

show abstract

Section: Ethical Implications Of Using Chatgpt and Ai Technologiesmentioning

confidence: 99%

ChatGPT as an important tool in organizational management: A review of the literature

Ayinde,

Wibowo,

Ravuri

et al. 2023

Business Information Review

View full text Add to dashboard Cite

show abstract

“…The growing usage of federated learning has prompted research into different types of attacks, such as backdoor attacks, 16,17 gradient leakage attacks, [18][19][20] poisoning attacks [21][22][23] and membership inference attacks. 24 The type of attack that is particularly relevant to our research is poisoning attacks, which can be classified into two categories: model poisoning and data poisoning.…”

Section: Literature Reviewmentioning

confidence: 99%

“…Moreover, these defenses are designed for model poisoning attacks. In contrast, in respect to data poisoning attacks, authors 22 proposed a class-wise cluster-based representation to detect the malicious participants attacked by data poisoning, and some other researchers 12,23 pointed out the effect of data poisoning on the performance of the global model. However, the research gap lies in the prevention strategy of data poisoning attacks that can effectively prevent such attacks.…”

Section: Literature Reviewmentioning

confidence: 99%

“…[3][4][5] With the rapid increase in computational capacity of edge devices [6][7][8] and advancements in efficient communication technologies with low power consumption, 9 FL is being applied nowadays in many fields. 10,11 The concept of federated machine learning involves collaborative machine learning model building across multiple workers while ensuring data privacy, where the raw data remains locally across numerous local devices. Despite recent progress in federated learning on privacy preservation, FL is still vulnerable to poisoning attacks, e.g., data poisoning attacks.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Confident federated learning to tackle label flipped data poisoning attacks

Ovi

Gangopadhyay

Erbacher

et al. 2023

Artificial Intelligence and Machine Learning for Multi-Domain Operations Applications V

View full text Add to dashboard Cite

Federated Learning (FL) enables collaborative model building among a large number of participants without revealing the sensitive data to the central server. However, because of its distributed nature, FL has limited control over the local data and corresponding training process. Therefore, it is susceptible to data poisoning attacks where malicious workers use malicious training data to train the model. Furthermore, attackers on the worker side can easily manipulate local data by swapping the labels of training instances to initiate data poisoning attacks. And local workers under such attacks carry incorrect information to the server, poison the global model, and cause misclassifications. So, detecting and preventing poisonous training samples from local training is crucial in federated training. To address it, we propose a federated learning framework, namely Confident Federated Learning to prevent data poisoning attacks on local workers. Here, we first validate the label quality of training samples by characterizing and identifying label errors in the training data and then exclude the detected mislabeled samples from the local training. To this aim, we experiment with our proposed approach on MNIST, Fashion-MNIST, and CIFAR-10 dataset and experimental results validated the robustness of the proposed framework against the data poisoning attacks by successfully detecting the mislabeled samples with above 85% accuracy.

show abstract

“…With the advent of federated learning, research in this domain has received a significant acceleration. FL is mainly utilized in the areas where data confidentiality and information security is of the utmost importance e.g., in contested environment [2]. By design, the worker nodes (e.g., mobile devices) in federated learning are protected from data intrusion; they are able to keep their original data private on their devices while training a model in alongside with one another, and they only need to send their local model updates to the server.…”

Section: Introductionmentioning

confidence: 99%

A Comprehensive Study of Gradient Inversion Attacks in Federated Learning and Baseline Defense Strategies

Ovi

Gangopadhyay

2023

2023 57th Annual Conference on Information Sciences and Systems (CISS)

Self Cite

View full text Add to dashboard Cite

show abstract

Towards developing a data security aware federated training framework in multi-modal contested environments

Cited by 11 publications

References 17 publications

ChatGPT as an important tool in organizational management: A review of the literature

ChatGPT as an important tool in organizational management: A review of the literature

Confident federated learning to tackle label flipped data poisoning attacks

A Comprehensive Study of Gradient Inversion Attacks in Federated Learning and Baseline Defense Strategies

Contact Info

Product

Resources

About