An Innovative Method for Mobile Network Payment Security System Based on hybrid LSTM-Ensemble Learning-Based Model
N Sujatha, S Pramod, S. Aravindh, B. Ramana Babu, S. Devi, M. Ananthi · 2024
It is common practice in many industries to use cash since it accounts for more than 85% of payments in almost all developing countries. A cell phone is become an everyday item. Mobile phones have practically become people’s constant companions due to the plethora of other uses they offer beyond just communicating. Since they are both cheap and multipurpose, soon everyone will be utilizing them. In an ideal world, everyone would use their cell phone for everything. Verification, preprocessing, and training the model are the three parts that make up the method. In order to confirm the payment, the customer’s biometric template is checked during the verification procedure. The three stages of fingerprint preparation include matching fingerprint templates, extracting fingerprint minutiae, and fingerprint preprocessing. The training of the model was carried out using a Hybrid LSTM-Ensemble Learning. The average accuracy of the suggested method is 92.76%, which is higher than LSTM and Ensemble Learning.