Deep learning technology is rapidly spreading in recent years and has been extensive attempts in the field of Brain-Computer Interface (BCI). Though the accuracy of Motor Imagery (MI) BCI systems based on the deep learning have been greatly improved compared with some traditional algorithms, it is still a big problem to clearly interpret the deep learning models. To address the issues, this work first introduces a popular deep learning model EEGNet and compares it with the traditional algorithm Filter-Bank Common Spatial Pattern (FBCSP). After that, this work considers that the 1-D convolution of EEGNet can be explained by a special Discrete Wavelet Transform (DWT), and the depthwise convolution of EEGNet is similar to the Common Spatial Pattern (CSP) algorithm. Therefore, this work improves the EEGNet by using the algorithm Temporary Constrained Sparse Group Lasso (TCSGL) to enhance its performance. The proposed model TSGL-EEGNet is tested on the BCI Competition IV 2a and BCI Competition III IIIa datasets that both are 4-classes classification MI tasks. The testing results show that the proposed model has achieved 78.96% (0.7194) average classification accuracy (kappa) on the dataset BCI Competition IV 2a, which are greater than EEGNet, C2CM, MB3DCNN, SS-MEMDBF and FBCSP, especially on insensitive subjects. The proposed model has also achieved 85.30% (0.8040) average classification accuracy (kappa) on the dataset BCI Competition III IIIa, which are greater than the EEGNet, MFTFS et al. At last, this work uses average-validation and stacking to further enhance the effect of the model. The 4-classes classification average accuracy rates reach 81.34% and 88.89%, and the kappas reach 0.7511 and 0.8519 on dataset BCI Competition IV 2a and BCI Competition III IIIa, respectively. Additionally, this work also uses the Grad-CAM to visualize the frequency and spatial features that are learned by the neural network.

A number of geodynamic models have been proposed for the southeastern margin of the Tibetan plateau and include a range of deformation processes. One unresolved issue is whether crustal and mantle flow occurs, and if it does, how flow contributes to the mass balance of the India‐Asia collision. To address this question, new magnetotelluric data were collected and used to derive a three‐dimensional electrical resistivity model of the crust and upper mantle beneath the Red River Fault (RRF) zone and adjacent areas. The most prominent features of the model are (1) a resistor in the upper‐middle crust directly beneath the trace of the RRF; (2) a major change in upper mantle resistivity across the RRF; and (3) a significant conductor in the upper mantle northeast of the RRF, which extends upward into the crust, and which requires a melt fraction of up to 3%. The model suggests that the lower crustal conductors may be due to melt/fluids derived from the mantle, rather than from outward flow from Tibet. The most likely source of fluids and melts could be upwelling mantle flow related to the Hainan mantle plume. The change in resistivity across the RRF implies a change in lithospheric strength may explain the present‐day localization of deformation and uplift at this location. The resistivity model may also give insights into the distribution of ore deposits in Ailao Shan, since many mineral deposits are derived from magmatic fluids generated in the mantle at the edges of regions of thick lithosphere.

Neural networks (NNs) are efficient tools for rapidly obtaining geoelectric models to solve magnetotelluric (MT) inversion problems. Training an NN with strong predictive power requires numerous training samples to prevent underfitting. To reduce the computational burden of generating a large number of training samples, this work analyzes the influence of the sample features and distribution on the training effect for an NN and proposes an efficient method of sample generation. This innovative method consists of three steps: 1) geoelectrically simplifying the features; 2) removing unnecessary features on the basis of realistic geological characteristics; and 3) mapping the samples to a higher-dimensional space. Numerical examples based on simple stratified models show that the number of samples can be reduced to below one-millionth of the original number while improving the predictive effect of the NN. The performance and effectiveness for processing more complex structures are verified by the inversion results obtained for a public data set, COPROD2. We conclude that this advanced method can generate high-quality training samples at a greatly reduced computational cost. The analysis of the sample features and distribution not only advances the state of research on the use of machine learning in geophysical inversion but also is a forward-looking study on the mechanisms of underfitting, tracing the source of these phenomena back to the training samples used.

As a trenchless detection approach, the transient electromagnetic method (TEM) can effectively detect well-conducted geoelectric structures, such as groundwater structures. The nonlinear TEM inversion process represented by neural network (NN) inversion does not rely on the initial model, which helps it efficiently and accurately obtain geoelectric structures. In this article, a factor analysis (FA)-based deep belief network (DBN) inversion framework built on hybrid memetic pretrained (HMP), HMP-FADBN, is proposed. We construct a DBN with restricted Boltzmann machines (RBMs) and backpropagation NNs (BPNNs) as the base fitting skeleton. FA is used to reduce the dimensionality of the feature space of the DBN. The hybrid memetic (HM) whale optimization algorithm (WOA) pretrains the network parameters and uses the memetic strategy to adjust the coordinated development system of social and individual cognition to enhance the network training effect. Numerical examples show that the prediction accuracy of the proposed HMP-FADBN for TEM geoelectric models is improved from more than 10% to approximately 2%. Moreover, 5%, 10%, and 15% noise tests show that the trained NN has good generalization and denoising abilities, and the prediction accuracy is less than 3% (affected by the maximum noise). Finally, the developed method is successfully applied to a landslide TEM survey, and the predicted quasi-2D geoelectric structure is consistent with the original geological structure.