SPECIAL ISSUE TITLE

Federated Learning For Collaborative Fraud Detection In Large-Scale Networks

Summary:

Throughout recent times, fraud involving credit cards has prompted institutions and customers to generate a great deal of revenue. In order to reduce the loss for businesses and customers, an efficient Fraud Detection System (FDS) is essential. A viable option that could deal with networking expenses, data security, and isolation challenges and is attracting major interest is Federated Learning (FL). The most distinctive feature of FL is that each device's local training dataset stays on the device; rather, every device computes a local model modification on its own and sends the information to a single server, which then combines all of the customers regional updates to develop the model globally constantly. Fraud involving credit cards has grown increasingly common in today's generation, and with the rise in internet crimes, numerous instances have already been documented in earlier times. While there are many methods available for identifying credit card fraud occurring online, completely online and federated learning approaches are particularly effective in identifying fraudulent activity.

The analysis indicates there are many more instances of fraudulent transactions than authorized activities in the charge card transaction dataset, which is highly biassed. Furthermore, sharing transaction information across banks is typically prohibited owing to data security and privacy concerns. Due to these issues, FDS finds it challenging to identify and understand trends in fraud. A delay in processing a credit card payment online could result in an unbearable expense and open an invitation to scammers. Therefore, it is crucial to have an online FDS that can handle time constraints and is capable of quickly identifying fraudulent activity. It is advisable to think about developing an effective fraud detection system that operates quickly enough to be applied in a real-time setting. Innovative structures are becoming more habitable, cost-effective, and sustainable due to the widespread use of Internet of Things (IoT) sensors. These sensors pick up environmental information and provide multidimensional temporal data that is critical for identifying irregularities and enhancing predictive utilization forecasting. However, a significant reaction-time lag frequently prevents these abnormalities from being detected in centralized systems. In order to get around this obstacle, the fraud detection challenge in a federated learning environment using the multitasking learning paradigm leverages patterns and disparities through tasks to solve numerous tasks at once.

The special issue employs several unsupervised learning algorithms, including an automated encoder, deployed over a federated learning framework to anticipate the extent of fraudulent credit card users. The number of fraudulent users is determined using transactions from a charge card dataset. The anarchic federated learning system is compared with the centralized technique.

List of Interested topics include, but are not limited to, the following:

A federated learning-oriented technique for financing fraud detection.

Bank fraud detection employing a federated learning model and data maintaining approaches.

Determining Data Fraud Detection using Collaborative Visualisation in Federated Learning.

An Intelligent Network-Based Approach to Identifying Fraud in Compensation Claims.

A Perceptive Network-Oriented Method for Recognising Fraud in Employment Allegations.

Potentially private encrypted multi-party computing for banking services using federated learning.

Moving towards federated network analysis for cooperative detection of financial offences.

Safeguarding Privacy while Identifying Financial Anomalies with Federated Learning.

Federated learning-based cooperative a pattern detection for the worldwide web of things.

A decentralised federated learning system with commission approval based on cryptocurrency.

Federated-learning oriented cryptocurrency IoMT system for medical with fraud-enabled privacy protection.

Special Issue Guest Editors:

Dr. Mahmud Iwan Solihin

Associate Professor at Faculty of Engineering,

UCSI University, Malaysia.

Email id: mahmudis@ucsiuniversity.edu.my, mahmud.iwan@proton.me

Google Scholar Link:

https://scholar.google.com.my/citations?user=BlBZJPUAAAAJ

Dr. Lin Guoping

Professor of Department of Engineering,

Department of Industrial Engineering and Enterprise Information,

Tunghai University, Taiwan.

Email id: kplin@thu.edu.tw

Google Scholar Link:

https://scholar.google.co.uk/citations?user=OyWUYLAAAAAJ

Dr. Slamet Riyadi

Associate Professor, Department of Information Technology,

Universitas Muhammadiyah Yogyakarta, Indonesia.

Email id: riyadi@umy.ac.id

Google Scholar Link:

https://scholar.google.com/citations?user=bl1BHx8AAAAJ&hl=en

Deadlines:

Article Submission Deadline - [10.08.2024]

Authors Notification Date - [20.10.2024]

Revised Papers Due Date - [25.12.2024]

Final notification Date - [20.02.2025]