|
|
 |
| |
|
|
|
Optimized Deep Fuzzy Neural Network for Financial Risk Evaluation in Fintech Model |
|
|
|
PP: 1507-1517 |
|
|
|
doi:10.18576/amis/180626
|
|
|
|
Author(s) |
|
|
|
Abdelsamie Eltayeb Tayfor,
Nadia Bushra Mohammed Ali,
Ibrahim Omer Elfaki,
|
|
|
|
Abstract |
|
|
| FinTech has evolved and invented rapidly in recent years. It uses a variety of services and technologies to improve economic processes, disrupt banks, and offer new solutions to clients and enterprises. Today, fintech refers to technology companies that directly serve consumers via mobile and online platforms, bypassing traditional financial institutions. The importance of banks in the economic model requires recognition of FinTech’s opportunities and threats for banks and their main roles as economic intermediaries in present economic services systems. Big data analytics, machine learning (ML), and artificial intelligence (AI) can improve risk assessment by providing real-time supplier economic health perceptions, influencing loan decisions and reducing defaults. This study proposes an Intelligent Multi-Criteria Decision Making Using a Deep Fuzzy Neural Network on Financial Risk Evaluation Enabled Fintech Model. The MDMDFNN-FREFM strategy uses deep learning and optimization to improve financial risk management decision-making. The MDMDFNN-FREFM method initially normalizes input data within a range using min-max normalization. To accurately quantify risk, the feature selection-based snake optimization (SO) model is used to determine the most important variables. Additionally, the deep fuzzy neural network (DFNN) model predicts financial risk. Finally, the dung beetle optimization (DBO) model is used to tune the DFNN model’s hyperparameters to reduce prediction errors and enhance accuracy. To improve MDMDFNN-FREFM performance, many simulation analyses are used. Under German credit and Polish companies bankruptcy datasets, MDMDFNN-FREFM performed well with 96.03% and 99.38% accuracy.
|
|
|
|
|
|
|
|
