The importance of parameter configuration in ECG-based biometric recognition models

This study examines the efficacy of an ECG-based biometric recognition model, concentrating specifically on the influence of iteration configurations and the quantity of nodes within the neural network. Through a series of experiments, we evaluate the false rejection rate (FRR) and false acceptance rate (FAR) in order to assess the model's efficacy.The results indicate that augmenting the number of nodes in the neural network markedly reduces the false rejection rate, decreasing from 0.5 (with 5 nodes) to 0.15 (with 15 nodes), hence demonstrating an improved capacity of the model to differentiate between individual ECG signals. On the other hand, the number of iterations, assessed from 100 to 800, had negligible impact on performance, with the FRR remaining stable and exhibiting just a tiny reduction as iterations increased. These findings underline the relevance of parameter setting in ECG-based biometric recognition models. While an enhanced number of nodes proved effective in decreasing false rejections, the impact of iteration count in model optimization needs further exploration. Future research directions should investigate the possibility for parameter modifications to maximize the performance and reliability of ECG-based biometric systems.