Optimisation of Attentional EEG Signal Classification Model Based on Attentional Mechanism of CNN-LSTM Algorithm

Deep learning has been widely applied to EEG signal classification in recent years, and this study aims to optimise and validate the effectiveness and performance advantages of an EEG signal classification model incorporating convolutional neural networks (CNNs), long and short-term memory networks (LSTMs), and attentional mechanisms in the field of attention classification . Multiple EEG signal data covering different attention levels are collected as experimental datasets. In the experimental preparation stage, the data are preprocessed, the sampling frequency and filtering parameters are set, and the dataset is divided into 70% training set, 15% validation set, and 15% test set. For model construction, CNN with TCN is used to extract the spatial features of EEG signals, LSTM captures the long-term dependencies in the time series, and the attention mechanism dynamically focuses on the part of the features that have a greater impact on the classification decision. The accuracy, recall and F1 value of the model in the EEG signal classification task are evaluated in the test set, and it is found that the CNN-LSTM-Attention model with TCN performs excellently in the task in this study, and is able to process the data and classify them accurately in a more efficient way, which can provide a valuable reference for the EEG signal related research and application.