Landslide Susceptibility Prediction Using Particle-Swarm-Optimized Multilayer Perceptron: Comparisons with Multilayer-Perceptron-Only, BP Neural Network, and Information Value Models

Li, Deying; Huang, Faming; Yan, Liangxuan; Cao, Zhongshan; Chen, Jiawu; Zhou, Ye

doi:10.3390/app9183664

Cited by 88 publications

(40 citation statements)

References 63 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The Multi-Layer Perceptron (MLP) constitutes the most conventional Artificial Neural Network (ANN) architecture [54]. It is formed by a set of simple elements called computational nodes or neurons, where each node sends a signal that depends on its activation state [55].…”

Section: Deep Learning Multi-layer Perceptronmentioning

confidence: 99%

“…Traditionally, MLP has been trained with the back-propagation algorithm (which is based in the stochastic gradient descent) and its weights randomly initialized [54,57]. However, in the latest versions of DL-MLPs, the hidden layers are pretrained by an unsupervised algorithm and the weights are optimized by the back-propagation algorithm [7].…”

Section: Deep Learning Multi-layer Perceptronmentioning

confidence: 99%

See 1 more Smart Citation

Data Sampling Methods to Deal With the Big Data Multi-Class Imbalance Problem

et al. 2020

View full text Add to dashboard Cite

The class imbalance problem has been a hot topic in the machine learning community in recent years. Nowadays, in the time of big data and deep learning, this problem remains in force. Much work has been performed to deal to the class imbalance problem, the random sampling methods (over and under sampling) being the most widely employed approaches. Moreover, sophisticated sampling methods have been developed, including the Synthetic Minority Over-sampling Technique (SMOTE), and also they have been combined with cleaning techniques such as Editing Nearest Neighbor or Tomek's Links (SMOTE+ENN and SMOTE+TL, respectively). In the big data context, it is noticeable that the class imbalance problem has been addressed by adaptation of traditional techniques, relatively ignoring intelligent approaches. Thus, the capabilities and possibilities of heuristic sampling methods on deep learning neural networks in big data domain are analyzed in this work, and the cleaning strategies are particularly analyzed. This study is developed on big data, multi-class imbalanced datasets obtained from hyper-spectral remote sensing images. The effectiveness of a hybrid approach on these datasets is analyzed, in which the dataset is cleaned by SMOTE followed by the training of an Artificial Neural Network (ANN) with those data, while the neural network output noise is processed with ENN to eliminate output noise; after that, the ANN is trained again with the resultant dataset. Obtained results suggest that best classification outcome is achieved when the cleaning strategies are applied on an ANN output instead of input feature space only. Consequently, the need to consider the classifier's nature when the classical class imbalance approaches are adapted in deep learning and big data scenarios is clear.

show abstract

Section: Deep Learning Multi-layer Perceptronmentioning

confidence: 99%

Section: Deep Learning Multi-layer Perceptronmentioning

confidence: 99%

Data Sampling Methods to Deal With the Big Data Multi-Class Imbalance Problem

et al. 2020

View full text Add to dashboard Cite

show abstract

“…In this study, the original spatial resolutions of the grid cells of the DEM and Landsat TM images were both 30 m. This resolution value can effectively characterize the topography of Shicheng County and can avoid excessive computation [73]. In addition, a lot of literature suggests that a spatial resolution of 30 m is feasible and satisfactory for LSP [6,30,48,70,74,75]. Hence, the spatial resolution of grid cells in this study was set to 30 m.…”

Section: Landslide-related Environmental Factorsmentioning

confidence: 99%

“…As such, this approach produces a higher prediction accuracy, can more precisely identify the nonlinear relationship between input and output variables, and retains more characteristic information from the original data [21][22][23][24]. This approach includes multiple adaptive regression splines [25,26], fuzzy logic [27,28], artificial neural network [15,29], multilayer perceptron [30], decision tree [31][32][33], random forest [34][35][36], support vector machine [37][38][39], rule-based approach [40], and multi-criteria evaluation techniques [41], among others.…”

Section: Introductionmentioning

confidence: 99%

Landslide Susceptibility Prediction Modeling Based on Remote Sensing and a Novel Deep Learning Algorithm of a Cascade-Parallel Recurrent Neural Network

Zhu

Huang

Fan

et al. 2020

Sensors

Self Cite

View full text Add to dashboard Cite

Landslide susceptibility prediction (LSP) modeling is an important and challenging problem. Landslide features are generally uncorrelated or nonlinearly correlated, resulting in limited LSP performance when leveraging conventional machine learning models. In this study, a deep-learning-based model using the long short-term memory (LSTM) recurrent neural network and conditional random field (CRF) in cascade-parallel form was proposed for making LSPs based on remote sensing (RS) images and a geographic information system (GIS). The RS images are the main data sources of landslide-related environmental factors, and a GIS is used to analyze, store, and display spatial big data. The cascade-parallel LSTM-CRF consists of frequency ratio values of environmental factors in the input layers, cascade-parallel LSTM for feature extraction in the hidden layers, and cascade-parallel full connection for classification and CRF for landslide/non-landslide state modeling in the output layers. The cascade-parallel form of LSTM can extract features from different layers and merge them into concrete features. The CRF is used to calculate the energy relationship between two grid points, and the extracted features are further smoothed and optimized. As a case study, the cascade-parallel LSTM-CRF was applied to Shicheng County of Jiangxi Province in China. A total of 2709 landslide grid cells were recorded and 2709 non-landslide grid cells were randomly selected from the study area. The results show that, compared with existing main traditional machine learning algorithms, such as multilayer perception, logistic regression, and decision tree, the proposed cascade-parallel LSTM-CRF had a higher landslide prediction rate (positive predictive rate: 72.44%, negative predictive rate: 80%, total predictive rate: 75.67%). In conclusion, the proposed cascade-parallel LSTM-CRF is a novel data-driven deep learning model that overcomes the limitations of traditional machine learning algorithms and achieves promising results for making LSPs.

show abstract

“…These models include some statistical models, such as the information value model (Li et al 2019), hierarchical clustering (Sheth et al 2001), the conditional probability model and the logistical regression model (Mousavi et al 2011;Papadopoulou-Vrynioti et al 2013;Sun et al 2017). In recent years, in addition to these statistical models, a few machine learning models have also been applied to carry out CSA, including the fuzzy mathematic method (Srivastava et al 2010;He et al 2013) and back-propagation neural network (Yilmaz et al 2013;Chen et al 2017;Li et al 2019). These machine learning models are considered to have higher prediction performance than that of statistical models because of their nonlinear fitting and prediction abilities (Yilmaz et al 2013;Huang et al 2017b;Wei et al 2018).…”

Section: Introductionmentioning

confidence: 99%

Collapse susceptibility assessment using a support vector machine compared with back-propagation and radial basis function neural networks

Sheng

Chai

et al. 2020

Geomatics, Natural Hazards and Risk

View full text Add to dashboard Cite

Machine learning models are regarded as efficient and popular models for natural disaster susceptibility prediction. However, few studies have focussed on the applications of the latest popular machine learning models in collapse susceptibility assessment (CSA). This paper proposes a 3S (RS, GPS and GIS) technologybased support vector machine (SVM) to map collapse susceptibility in the Nantian area of China. The 3S technology-based backpropagation neural network (BPNN) and radial basis function neural network (RBFNN) models are also proposed for comparison. First, 44 recorded collapses are identified through field investigation, and fourteen collapse-related causal factors are acquired using 3S technology. Second, among these recorded collapses and randomly selected 'non-collapses', 70% of the collapse and non-collapse grid cells are used to train the three models, while the remaining 30% are used to test the models. Third, the collapse susceptibility maps of the Nantian area are produced using the three models. Finally, the prediction accuracies of these models are evaluated. The results indicate that the SVM model has the highest prediction accuracy, while the RBFNN model has the lowest prediction accuracy for CSA. In addition, the distribution characteristics of collapse susceptibility in the Nantian area are produced well by all three models. ARTICLE HISTORY

show abstract

Landslide Susceptibility Prediction Using Particle-Swarm-Optimized Multilayer Perceptron: Comparisons with Multilayer-Perceptron-Only, BP Neural Network, and Information Value Models

Cited by 88 publications

References 63 publications

Data Sampling Methods to Deal With the Big Data Multi-Class Imbalance Problem

Data Sampling Methods to Deal With the Big Data Multi-Class Imbalance Problem

Landslide Susceptibility Prediction Modeling Based on Remote Sensing and a Novel Deep Learning Algorithm of a Cascade-Parallel Recurrent Neural Network

Collapse susceptibility assessment using a support vector machine compared with back-propagation and radial basis function neural networks

Contact Info

Product

Resources

About