Combining Evolutionary Algorithms and Machine Learning Models in Landslide Susceptibility Assessments

The main objective of the present study is to introduce a novel predictive model that combines evolutionary algorithms and machine learning (ML) models, so as to construct a landslide susceptibility map. Genetic algorithms (GA) are used as a feature selection method, whereas the particle swarm optimization (PSO) method is used to optimize the structural parameters of two ML models, support vector machines (SVM) and artificial neural network (ANN). A well-defined spatial database, which included 335 landslides and twelve landslide-related variables (elevation, slope angle, slope aspect, curvature, plan curvature, profile curvature, topographic wetness index, stream power index, distance to faults, distance to river, lithology, and hydrological cover) are considered for the analysis, in the Achaia Regional Unit located in Northern Peloponnese, Greece. The outcome of the study illustrates that both ML models have an excellent performance, with the SVM model achieving the highest learning accuracy (0.977 area under the receiver operating characteristic curve value (AUC)), followed by the ANN model (0.969). However, the ANN model shows the highest prediction accuracy (0.800 AUC), followed by the SVM (0.750 AUC) model. Overall, the proposed ML models highlights the necessity of feature selection and tuning procedures via evolutionary optimization algorithms and that such approaches could be successfully used for landslide susceptibility mapping as an alternative investigation tool.

Download Full-text

Performance Evaluation and Comparison of Bivariate Statistical-Based Artificial Intelligence Algorithms for Spatial Prediction of Landslides

ISPRS International Journal of Geo-Information ◽

10.3390/ijgi9120696 ◽

2020 ◽

Vol 9 (12) ◽

pp. 696

Author(s):

Wei Chen ◽

Zenghui Sun ◽

Xia Zhao ◽

Xinxiang Lei ◽

Ataollah Shirzadi ◽

...

Keyword(s):

Landslide Susceptibility ◽

Vegetation Index ◽

Slope Angle ◽

Spatial Prediction ◽

Slope Aspect ◽

Support Vector ◽

Topographic Wetness Index ◽

Bivariate Model ◽

Susceptibility Maps ◽

Landslide Susceptibility Maps

The purpose of this study is to compare nine models, composed of certainty factors (CFs), weights of evidence (WoE), evidential belief function (EBF) and two machine learning models, namely random forest (RF) and support vector machine (SVM). In the first step, fifteen landslide conditioning factors were selected to prepare thematic maps, including slope aspect, slope angle, elevation, stream power index (SPI), sediment transport index (STI), topographic wetness index (TWI), plan curvature, profile curvature, land use, normalized difference vegetation index (NDVI), soil, lithology, rainfall, distance to rivers and distance to roads. In the second step, 152 landslides were randomly divided into two groups at a ratio of 70/30 as the training and validation datasets. In the third step, the weights of the CF, WoE and EBF models for conditioning factor were calculated separately, and the weights were used to generate the landslide susceptibility maps. The weights of each bivariate model were substituted into the RF and SVM models, respectively, and six integrated models and landslide susceptibility maps were obtained. In the fourth step, the receiver operating characteristic (ROC) curve and related parameters were used for verification and comparison, and then the success rate curve and the prediction rate curves were used for re-analysis. The comprehensive results showed that the hybrid model is superior to the bivariate model, and all nine models have excellent performance. The WoE–RF model has the highest predictive ability (AUC_T: 0.9993, AUC_P: 0.8968). The landslide susceptibility maps produced in this study can be used to manage landslide hazard and risk in Linyou County and other similar areas.

Download Full-text

A Semi-Automated Object-Based Gully Networks Detection Using Different Machine Learning Models: A Case Study of Bowen Catchment, Queensland, Australia

Sensors ◽

10.3390/s19224893 ◽

2019 ◽

Vol 19 (22) ◽

pp. 4893 ◽

Cited By ~ 22

Author(s):

Hejar Shahabi ◽

Ben Jarihani ◽

Sepideh Tavakkoli Piralilou ◽

David Chittleborough ◽

Mohammadtaghi Avand ◽

...

Keyword(s):

Machine Learning ◽

Image Segmentation ◽

Vegetation Index ◽

Scale Parameter ◽

Slope Aspect ◽

Support Vector ◽

Topographic Wetness Index ◽

Slope Length ◽

Optimal Scale ◽

Object Based

Gully erosion is a dominant source of sediment and particulates to the Great Barrier Reef (GBR) World Heritage area. We selected the Bowen catchment, a tributary of the Burdekin Basin, as our area of study; the region is associated with a high density of gully networks. We aimed to use a semi-automated object-based gully networks detection process using a combination of multi-source and multi-scale remote sensing and ground-based data. An advanced approach was employed by integrating geographic object-based image analysis (GEOBIA) with current machine learning (ML) models. These included artificial neural networks (ANN), support vector machines (SVM), and random forests (RF), and an ensemble ML model of stacking to deal with the spatial scaling problem in gully networks detection. Spectral indices such as the normalized difference vegetation index (NDVI) and topographic conditioning factors, such as elevation, slope, aspect, topographic wetness index (TWI), slope length (SL), and curvature, were generated from Sentinel 2A images and the ALOS 12-m digital elevation model (DEM), respectively. For image segmentation, the ESP2 tool was used to obtain three optimal scale factors. On using object pureness index (OPI), object matching index (OMI), and object fitness index (OFI), the accuracy of each scale in image segmentation was evaluated. The scale parameter of 45 with OFI of 0.94, which is a combination of OPI and OMI indices, proved to be the optimal scale parameter for image segmentation. Furthermore, segmented objects based on scale 45 were overlaid with 70% and 30% of a prepared gully inventory map to select the ML models’ training and testing objects, respectively. The quantitative accuracy assessment methods of Precision, Recall, and an F1 measure were used to evaluate the model’s performance. Integration of GEOBIA with the stacking model using a scale of 45 resulted in the highest accuracy in detection of gully networks with an F1 measure value of 0.89. Here, we conclude that the adoption of optimal scale object definition in the GEOBIA and application of the ensemble stacking of ML models resulted in higher accuracy in the detection of gully networks.

Download Full-text

Landslides susceptibility modelling using Multivariate Logistic Regression Model in the Sahla Watershed in Northern Morocco

Sociedade & Natureza ◽

10.14393/sn-v33-2021-59124 ◽

2021 ◽

Vol 33 ◽

Author(s):

Mohammed El-Fengour ◽

Hanifa El Motaki ◽

Aissa El Bouzidi

Keyword(s):

Logistic Regression ◽

Landslide Susceptibility ◽

Slope Angle ◽

Google Earth ◽

Slope Aspect ◽

Multivariate Logistic Regression Model ◽

Susceptibility Map ◽

Potential Hazard ◽

Topographic Wetness Index ◽

Landslide Occurrence

This study aimed to assess landslide susceptibility in the Sahla watershed in northern Morocco. Landslides hazard is the most frequent phenomenon in this part of the state due to its mountainous precarious environment. The abundance of rainfall makes this area suffer mass movements led to a notable adverse impact on the nearby settlements and infrastructures. There were 93 identified landslide scars. Landslide inventories were collected from Google Earth image interpretations. They were prepared out of landslide events in the past, and future landslide occurrence was predicted by correlating landslide predisposing factors. In this paper, landslide inventories are divided into two groups, one for landslide training and the other for validation. The Landslide Susceptibility Map (LSM) is prepared by Logistic Regression (LR) Statistical Method. Lithology, stream density, land use, slope curvature, elevation, topographic wetness index, slope aspect, and slope angle were used as conditioning factors. The Area Under the Curve (AUC) of the Receiver Operating Characteristic (ROC) was employed to examine the performance of the model. In the analysis, the LR model results in 96% accuracy in the AUC. The LSM consists of the predicted landslide area. Hence it can be used to reduce the potential hazard linked with the landslides in the Sahla watershed area in Rif Mountains in northern Morocco.

Download Full-text

Landslide Susceptibility Mapping using Bivariate Statistical Models and GIS in Chattagram District, Bangladesh

10.21203/rs.3.rs-531858/v1 ◽

2021 ◽

Author(s):

Md. Sharafat Chowdhury ◽

Bibi Hafsa

Keyword(s):

Statistical Models ◽

Landslide Susceptibility ◽

Slope Angle ◽

Information Value ◽

Weight Of Evidence ◽

Landslide Susceptibility Mapping ◽

Slope Aspect ◽

Susceptibility Map ◽

Topographic Wetness Index ◽

Conditioning Factors

Abstract This study attempts to produce Landslide Susceptibility Map for Chattagram District of Bangladesh by using five GIS based bivariate statistical models, namely the Frequency Ratio (FR), Shanon’s Entropy (SE), Weight of Evidence (WofE), Information Value (IV) and Certainty Factor (CF). A secondary landslide inventory database was used to correlate the previous landslides with the landslide conditioning factors. Sixteen landslide conditioning factors of Slope Aspect, Slope Angle, Geology, Elevation, Plan Curvature, Profile Curvature, General Curvature, Topographic Wetness Index, Stream Power Index, Sediment Transport Index, Topographic Roughness Index, Distance to Stream, Distance to Anticline, Distance to Fault, Distance to Road and NDVI were used. The Area Under Curve (AUC) was used for validation of the LSMs. The predictive rate of AUC for FR, SE, WofE, IV and CF were 76.11%, 70.11%, 78.93%, 76.57% and 80.43% respectively. CF model indicates 15.04% of areas are highly susceptible to landslide. All the models showed that the high elevated areas are more susceptible to landslide where the low-lying river basin areas have a low probability of landslide occurrence. The findings of this research will contribute to land use planning, management and hazard mitigation of the CHT region.

Download Full-text

A Comparative Study of Landslide Susceptibility Mapping Using SVM and PSO-SVM Models Based on Grid and Slope Units

Mathematical Problems in Engineering ◽

10.1155/2021/8854606 ◽

2021 ◽

Vol 2021 ◽

pp. 1-15

Author(s):

Shuai Zhao ◽

Zhou Zhao

Keyword(s):

Support Vector Machine ◽

Landslide Susceptibility ◽

Learning Algorithm ◽

Slope Angle ◽

Susceptibility Mapping ◽

Landslide Susceptibility Mapping ◽

Training Dataset ◽

Validation Dataset ◽

Slope Aspect ◽

Support Vector

The main purpose of this study aims to apply and compare the rationality of landslide susceptibility maps using support vector machine (SVM) and particle swarm optimization coupled with support vector machine (PSO-SVM) models in Lueyang County, China, enhance the connection with the natural terrain, and analyze the application of grid units and slope units. A total of 186 landslide locations were identified by earlier reports and field surveys. The landslide inventory was randomly divided into two parts: 70% for training dataset and 30% for validation dataset. Based on the multisource data and geological environment, 16 landslide conditioning factors were selected, including control factors and triggering factors (i.e., altitude, slope angle, slope aspect, plan curvature, profile curvature, SPI, TPI, TRI, lithology, distance to faults, TWI, distance to rivers, NDVI, distance to roads, land use, and rainfall). The susceptibility between each conditioning factor and landslide was deduced using a certainty factor model. Subsequently, combined with grid units and slope units, the landslide susceptibility models were carried out by using SVM and PSO-SVM methods. The precision capability of the landslide susceptibility mapping produced by different models and units was verified through a receiver operating characteristic (ROC) curve. The results showed that the PSO-SVM model based on slope units had the best performance in landslide susceptibility mapping, and the area under the curve (AUC) values of training and validation datasets are 0.945 and 0.9245, respectively. Hence, the machine learning algorithm coupled with slope units can be considered a reliable and effective technique in landslide susceptibility mapping.

Download Full-text

Landslide susceptibility assessment and mapping using statistical and data mining models in Iran

10.21203/rs.3.rs-239985/v1 ◽

2021 ◽

Author(s):

Abdulaziz Hanifinia ◽

Habib Nazarnejad ◽

Saeed Najafi ◽

Aiding Kornejady ◽

Hamid Reza Pourghasemi

Keyword(s):

Data Mining ◽

Land Use ◽

Shannon Entropy ◽

Landslide Susceptibility ◽

Goodness Of Fit ◽

Slope Angle ◽

Google Earth ◽

Slope Aspect ◽

Support Vector ◽

The Village

Abstract This study applied to evaluate landslide susceptibility using four data mining models including, “Generalized Linear Model (GLM)”, “Maximum Entropy (ME)”, “Artificial Neural Network (ANN)”, and “Support Vector Machine (SVM)” in Cherikabad Watershed in Urmia City, Iran. In particular, Shannon entropy was used to assess the intercomparison of factors’ classes. Eleven factors including, elevation, slope angle, slope aspect, geological formation, annual mean rainfall, land use/ land cover, distance to the village, distance to faults, distance to roads, distance to streams, and NDVI used in the current study. Landslide inventory map was identified using Google Earth imagery, extensive field surveys, and scrutinizing archived data. The produced landslide susceptibility maps were evaluated by the AUROC index. The results of performance metrics revealed that the Shannon entropy with an AUROC of 0.879 proved highly reliable and so is the intercomparison analysis of factors’ classes derived from it. Additionally, the goodness-of-fit of the GLM, ME, ANN, and SVM models were 0.763, 0.740, 0.926, and 0.924, while their predictive powers were 0.751, 0.727, 0.917, and 0.935, respectively. Hence, the results indicated that the SVM model can be introduced as the superior model for the study area based on which the most critical factors affecting landslides were found to be elevation, annual mean rainfall, and distance to the village. The results of this work are of great use for land use planning in landslide-prone areas with similar geo-topological, geomorphological, and climatic conditions.

Download Full-text

Multi-Hazard Exposure Mapping Using Machine Learning for the State of Salzburg, Austria

Remote Sensing ◽

10.3390/rs12172757 ◽

2020 ◽

Vol 12 (17) ◽

pp. 2757 ◽

Cited By ~ 2

Author(s):

Thimmaiah Gudiyangada Nachappa ◽

Omid Ghorbanzadeh ◽

Khalil Gholamnia ◽

Thomas Blaschke

Keyword(s):

Machine Learning ◽

Vegetation Index ◽

Accuracy Assessment ◽

The State ◽

Receiver Operating Curve ◽

Slope Aspect ◽

Federal State ◽

Support Vector ◽

Topographic Wetness Index ◽

Hazard Exposure

We live in a sphere that has unpredictable and multifaceted landscapes that make the risk arising from several incidences that are omnipresent. Floods and landslides are widespread and recurring hazards occurring at an alarming rate in recent years. The importance of this study is to produce multi-hazard exposure maps for flooding and landslides for the federal State of Salzburg, Austria, using the selected machine learning (ML) approach of support vector machine (SVM) and random forest (RF). Multi-hazard exposure maps were established on thirteen influencing factors for flood and landslides such as elevation, slope, aspect, topographic wetness index (TWI), stream power index (SPI), normalized difference vegetation index (NDVI), geology, lithology, rainfall, land cover, distance to roads, distance to faults, and distance to drainage. We classified the inventory data for flood and landslide into training and validation with the widely used splitting ratio, where 70% of the locations are used for training, and 30% are used for validation. The accuracy assessment of the exposure maps was derived through ROC (receiver operating curve) and R-Index (relative density). RF yielded better results for both flood and landslide exposure with 0.87 for flood and 0.90 for landslides compared to 0.87 for flood and 0.89 for landslides using SVM. However, the multi-hazard exposure map for the State of Salzburg derived through RF and SVM provides the planners and managers to plan better for risk regions affected by both floods and landslides.

Download Full-text

Landslide Susceptibility Assessment Using Integrated Deep Learning Algorithm along the China-Nepal Highway

Sensors ◽

10.3390/s18124436 ◽

2018 ◽

Vol 18 (12) ◽

pp. 4436 ◽

Cited By ~ 24

Author(s):

Liming Xiao ◽

Yonghong Zhang ◽

Gongzhuang Peng

Keyword(s):

Landslide Susceptibility ◽

Vegetation Index ◽

Learning Algorithm ◽

Dynamic Change ◽

Slope Angle ◽

Machine Learning Algorithms ◽

Sensor Data ◽

Slope Aspect ◽

Support Vector ◽

Important Indicator

The China-Nepal Highway is a vital land route in the Kush-Himalayan region. The occurrence of mountain hazards in this area is a matter of serious concern. Thus, it is of great importance to perform hazard assessments in a more accurate and real-time way. Based on temporal and spatial sensor data, this study tries to use data-driven algorithms to predict landslide susceptibility. Ten landslide instability factors were prepared, including elevation, slope angle, slope aspect, plan curvature, vegetation index, built-up index, stream power, lithology, precipitation intensity, and cumulative precipitation index. Four machine learning algorithms, namely decision tree (DT), support vector machines (SVM), Back Propagation neural network (BPNN), and Long Short Term Memory (LSTM) are implemented, and their final prediction accuracies are compared. The experimental results showed that the prediction accuracies of BPNN, SVM, DT, and LSTM in the test areas are 62.0%, 72.9%, 60.4%, and 81.2%, respectively. LSTM outperformed the other three models due to its capability to learn time series with long temporal dependencies. It indicates that the dynamic change course of geological and geographic parameters is an important indicator in reflecting landslide susceptibility.

Download Full-text

Landslide Susceptibility Zonation Using Statistical and Machine Learning Approaches in Northern Lecco, Italy

10.21203/rs.3.rs-771789/v1 ◽

2021 ◽

Author(s):

Mohammad Mehrabi

Keyword(s):

Machine Learning ◽

Landslide Susceptibility ◽

Landslide Susceptibility Mapping ◽

Slope Aspect ◽

Learning Approaches ◽

Visual Interpretation ◽

Topographic Wetness Index ◽

Inference System ◽

Conditioning Factors ◽

Susceptibility Maps

Abstract This study deals with landslide susceptibility mapping in the northern part of Lecco Province, Lombardy Region, Italy. In so doing, a valid landslide inventory map and thirteen conditioning factors (including elevation, slope aspect, slope degree, plan curvature, profile curvature, distance to waterway, distance to road, distance to fault, soil type, land use, lithology, stream power index, and topographic wetness index) form the spatial database within geographic information system (GIS). The used evaluative models comprise a bivariate statistical approach called frequency ratio (FR) and two machine learning tools, namely multi-layer perceptron neural network (MLPNN) and adaptive neuro-fuzzy inference system (ANFIS). These models first use landslide and non-landslide records for comprehending the relationship between the landslide occurrence and conditioning factors. Then landslide susceptibility values are predicted for the whole area. The accuracy of the produced susceptibility maps is measured using area under the curve (AUC) index, according to which, the MLPNN (AUC = 0.916) presented the most accurate map, followed by the FR (AUC = 0. 898) and ANFIS (AUC = 0.889). Visual interpretation of the susceptibility maps, FR-based correlation analysis, as well as the importance assessment of conditioning factors, all indicated the significant contribution of the road networks to the crucial susceptibility of landslide. Lastly, an explicit predictive formula is extracted from the implemented MLPNN model for a convenient approximation of landslide susceptibility value.

Download Full-text

A Comparison of Feature Selection and Forecasting Machine Learning Algorithms for Predicting Glycaemia in Type 1 Diabetes Mellitus

Applied Sciences ◽

10.3390/app11041742 ◽

2021 ◽

Vol 11 (4) ◽

pp. 1742

Author(s):

Ignacio Rodríguez-Rodríguez ◽

José-Víctor Rodríguez ◽

Wai Lok Woo ◽

Bo Wei ◽

Domingo-Javier Pardo-Quiles

Keyword(s):

Diabetes Mellitus ◽

Machine Learning ◽

Type 1 Diabetes ◽

Feature Selection ◽

Blood Glucose ◽

Type 1 Diabetes Mellitus ◽

Support Vector ◽

Chronic Hyperglycemia ◽

Predictive Algorithms

Type 1 diabetes mellitus (DM1) is a metabolic disease derived from falls in pancreatic insulin production resulting in chronic hyperglycemia. DM1 subjects usually have to undertake a number of assessments of blood glucose levels every day, employing capillary glucometers for the monitoring of blood glucose dynamics. In recent years, advances in technology have allowed for the creation of revolutionary biosensors and continuous glucose monitoring (CGM) techniques. This has enabled the monitoring of a subject’s blood glucose level in real time. On the other hand, few attempts have been made to apply machine learning techniques to predicting glycaemia levels, but dealing with a database containing such a high level of variables is problematic. In this sense, to the best of the authors’ knowledge, the issues of proper feature selection (FS)—the stage before applying predictive algorithms—have not been subject to in-depth discussion and comparison in past research when it comes to forecasting glycaemia. Therefore, in order to assess how a proper FS stage could improve the accuracy of the glycaemia forecasted, this work has developed six FS techniques alongside four predictive algorithms, applying them to a full dataset of biomedical features related to glycaemia. These were harvested through a wide-ranging passive monitoring process involving 25 patients with DM1 in practical real-life scenarios. From the obtained results, we affirm that Random Forest (RF) as both predictive algorithm and FS strategy offers the best average performance (Root Median Square Error, RMSE = 18.54 mg/dL) throughout the 12 considered predictive horizons (up to 60 min in steps of 5 min), showing Support Vector Machines (SVM) to have the best accuracy as a forecasting algorithm when considering, in turn, the average of the six FS techniques applied (RMSE = 20.58 mg/dL).

Download Full-text