Electroencephalogram-Based Attention Level Classification Using Convolution Attention Memory Neural Network

Attentive learning is an important feature of the process. It provides a beneficial experience and plays key role in generating positive outcomes. Most studies widely applied electroencephalogram (EEG) to measure human attention level. Although most use EEG handcrafted features statistical methods classify level, more effective technique still needed. In this paper, we aim analyze participants’ signals through deep model those as showing either attentive or inattentive behaviors. To carry out research, initially conducted background study on its detection EEG. After that, design Troxler’s fading experiment device collect data behaviors during test. The collected will be analyzed using Convolution Attention Memory Neural Network (CAMNN) proposed CAMNN optimized with Vector-to-Vector (Vec2Vec) modeling, where can learned neural networks end-to-end approach. result shows that our achieve 92% accuracy 0.92 F1 score which outperforms several existing network models such Recurrent (RNN), Long Short-Term (LSTM), Convolutional (CNN), Deep Learning Networks (deep ConvNets), Compact for EEG-based BCIs (EEGNet). This research useful who are interested developing level monitoring biofeedback system areas educational classroom learning, medical industrial operator.