Performance Enhancement and Detection of Non Rapid Eye Movement using CNN Comparison over Recurrent Neural Network

Abstract

Aim: The main aim of this work is to detect novel non rapid eye movements using Convolution Neural Networks (CNN) and comparing them with Recurrent Neural Networks (RNN).

Materials and Methods: Two groups, namely Convolution Neural Network and Recurrent Neural Network algorithm were used to find the accuracy of Novel Non rapid eye movement detection with 15 samples each to evaluate, a total of 30 samples were taken for this study. Sample size was calculated using G power with pretest power at 80%, with the error rate of 0.05. Data models were trained using these Neural networks algorithms where the Novel Non rapid eye movement adapts the system effectively.

Results : The accuracy of the Convolution Neural Network is significantly (0.013; p<0.05) improved with 95.65% than the RNN with 92.1%.

Conclusion: CNN algorithm has obtained more accuracy which significantly shows differences when compared to the RNN algorithm.