Graph networks are naturally suitable for modeling multi-channel features of EEG signals. However, the existing study that attempts to utilize graph-based neural networks for EEG-based emotion recognition doesn’t take the spatio-temporal redundancy of EEG features and differences in brain topology into account. In this paper, we propose EEG-GCN, a paradigm that adopts spatio-temporal and self-adaptive graph convolutional networks for single and multi-view EEG-based emotion recognition. With spatio-temporal attention mechanism employed, EEG-GCN can adaptively capture significant sequential segments and spatial location information in EEG signals. Meanwhile, a self-adaptive brain network adjacency matrix is designed to quantify the connection strength between the channels, in which way to represent the diverse activation patterns under different emotion scenarios. Additionally, we propose a multi-view EEG-based emotion recognition method, which effectively integrates the diverse features of EEG signals. Extensive experiments conducted on two benchmark datasets SEED and DEAP demonstrate that our proposed method outperforms other representative methods from both single and multiple views.

With the development of Internet lending, the use of peer-to-peer (P2P) as a new financial credit model has been increasing in China. However, this rapid development has led to a major potential risk. A few P2P enterprises operate well in the beginning but close within a short period because of the suspension of business, fraud, illegal fundraising, and blind expansion. Effective supervision of the P2P industry is an urgent problem. Trading volume reflects the operation stability of P2P platforms. Hence, predicting the volume of the P2P market is an important research topic. This paper first analyzes the trading data of a P2P platform. It is found that the sentiment of investor comments is related to the trading volume of the P2P platform. Then, we use the TextCNN model to classify the sentiment of investor comments and obtain the time series of changes in sentiment. It is verified that the time series of change in sentiment and the P2P volume index has statistical causality and a strong correlation. This paper proposes a model that uses the trend of change in investor sentiment to predict P2P trading volume. This model uses the historical time series change in investor sentiment, the P2P volume index, and WeekDay characteristics to predict future P2P trading volume. The experimental results show that the proposed model is better than a few existing baseline methods. Compared with baseline regression, the Pearson coefficient of the predicted and actual values of the proposed model is increased by 13.26%, the mean squared error is decreased by 27.62%, and the R-squared value is increased by 28.48%.