Review on Adaptive Learning Framework for Cyber Financial Thread Detection
Main Article Content
Abstract
As digital financial services (DFS) have grown rapidly, they have become more vulnerable to advanced cyberattacks and require intelligent and adaptive security solutions. This review intends to explore the most recent developments in the Adaptive Learning frameworks for Cyber Financial Threat Detection, focusing on techniques, applications, performance analysis and future research directions. To analyze recent studies on the development of deep learning-based cybersecurity models, such as Convolutional Neural Networks (CNNs), Long Short-Term Memory (LSTM) networks, Autoencoders, Transformerbased architectures, and Graph Neural Networks (GNNs) were used in a systematic review approach. The review also covers popular benchmark datasets, evaluation methods and some new technologies like Explainable Artificial Intelligence (XAI) and Federated Learning. The results revealed that adaptive deep learning is effective in detecting fraud, intrusion, malware, phishing and anomaly than traditional rule-based detection methods. Detected studies were found to have high detection accuracy, higher precision and recall rate, lower false-positive rates, and better adaptability to new cyber threats. However, there are still some limitations in the area of computational complexity, small number of labeled data, each model's interpretability, and its deployment in real time. The review underscores the opportunities and challenges of adaptive deep learning to offer scalable, intelligent, and resilient cybersecurity solutions, and suggests research areas that could be fruitful for building a more secure digital financial ecosystem.
Article Details

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
References
E. Walker, S. Mitchell, B. Carter, O. Reynolds, and C. Paul, “Adaptive AI Frameworks for Financial Fraud and Cyber Threat Detection Authors”.
A. A. Hammad, S. R. Ahmed, and M. R. Ahmed, Deep Reinforcement Learning for Adaptive Cyber Defense in Network Security, vol. 1, no. 1. Association for Computing Machinery. doi: 10.1145/3660853.3660930.
A. Carlo, “Quantum Computing’ s Disruption of Cyber Security: Challenges and MLBased Solutions,” no. February, 2025, doi: 10.13140/RG.2.2.23148.68482.
E. Oye, “Blockchain-Based Systems For Secure Machine Learning In Cybersecurity,” SSRN Electron. J., no. December, 2025, doi: 10.2139/ssrn.5080340.
E. Oye, “Blockchain-Based Systems For Secure Machine Learning In Cybersecurity,” SSRN Electron. J., 2025, doi: 10.2139/ssrn.5080340.