Time series landslide displacement is the most critical data set to understand landslide characteristics and infer its future development. To predict landslide displacements and their quantitative uncertainties, a mathematical description of the landslide evolution should be established. This paper proposes a novel hybrid machine-learning model to predict landslide displacements and quantify their uncertainties using prediction intervals (PIs). First, wavelet de-noising and Hodrick-Prescott (HP) filters are applied to decompose the original landslide displacement into periodic, trend, and noise components. Second, a module built on the framework of bootstrap and extreme learning machine (ELM) with a hybrid grey wolf optimizer (HGWO) is used to derive a formula for modelling the periodic component of the landslide motion. Another formula for predicting the trend component of the displacement is derived by double exponential smoothing (DES). Grey relational analysis and kernel principal component analysis (KPCA) are used to select the main factors controlling the landslide motions. Finally, the two constructed formulas are used to generate the predictions of landslide displacements and the PIs. Validation and comparison experiments have been carried out on the Baishuihe landslide in the Three Gorge Reservoir of China. Results demonstrate the proposed method can achieve better performance with higher-quality PIs than other state-of-theart methods.

No abstract

Recently, the classification of hyperspectral images has made great process. Especially, the classification methods based on three-dimensional convolutional neural network have remarkable performance due to the uniqueness of hyperspectral images. However, the hyperspectral classification still faces great challenges due to a series of problems such as the insufficient extraction of spectral-spatial features, the lack of labeled samples, the large amount of noise, the tendency of overfitting and so on. Therefore, SSDANet is proposed to solve the above problems and promote the further development of hyperspectral classification technology based on deep learning. SSDANet is a spectral-spatial three-dimensional convolutional neural network with a deep and wide structure that can significantly improve classification performance. In SSDANet, the spectral-spatial dense connectivity is put forward to protect the integrity of information. It is made up of the spectral branch and the spatial branch, which can learn and reuse the spectralspatial features. Besides, the spectral-spatial attention mechanism is proposed to adapt the special structure of hyperspectral images. It can excite important spectral-spatial information and suppress unimportant spectral-spatial information. In addition, a series of optimization methods including data augmentation, batch normalization, dropout, exponential decay learning rate, and L2 regularization are adopted to alleviate the problem of overfitting and improve the classification results. To verify the performance of SSDANet, experiments were implemented on two widely used datasets-Pavia University and Indian Pines. Under the condition of limited labeled samples, the classification evaluation indexes of OA, AA, and Kappa on the two datasets all exceeded 99%, reaching state-of-the-art performance. INDEX TERMS artificial intelligence, hyperspectral imaging, image processing, pattern recognition, remote sensing