Search

Landslide susceptibility analysis can provide theoretical support for landslide risk management. However, some susceptibility analyses are not sufficiently interpretable. Moreover, the accuracy of many research methods needs to be improved. Therefore, this study can supplement these deficiencies. This study aims to research the evaluation effects of random forest (RF) and extreme gradient boosting (XGBoost) classifier models on landslide susceptibility, and to compare their applicability in Fengjie County, Chongqing, a typical landslide‐prone area in southwest of China. Firstly, 1624 landslides information from 1980 to 2020 were obtained through field investigation, and a geospatial database of 16 conditional factors had been constructed. Secondly, non‐landslide points were selected to form a complete data set and RF and XGBoost models were established. Finally, the area under the ROC curve (AUC) value, accuracy, and F‐score were used to compare the two models. The results show that even though both classifiers have a highly accurate evaluation of landslide susceptibility, the RF model performs better. In comparison, the RF model has a higher AUC value of 0.866, and its accuracy and F‐score are approximately 2% higher than XGBoost. The land use, elevation, and lithology of Fengjie County contribute to the occurrence of landslides. This is due to human engineering activities (such as land reclamation, and housing construction) resulting in low slope stability and landslides in widely distributed sandstone, siltstone, and mudstone layers owing to their low permeability and planes of weakness. [ABSTRACT FROM AUTHOR]

Electricity load forecasting is one of the important tasks of the power marketing department, and accurate load forecasting is extremely important to ensure real-time dispatch and security of the power system. In order to obtain accurate and reliable load forecasting results, an ultra-short-term power load forecasting model based on an improved random forest regression algorithm is proposed in this paper. First, data pre-processing is performed on the original dataset. Then the pre-processed time data and historical load data are used as inputs to the model, and optimization of the model using the Gaussian mixture-based tree-structured Parzen estimator algorithm is carried out. Finally, the final prediction results were derived. Experimental analysis was conducted with real load data from a region of China, and the experimental results show that the method has better prediction accuracy than the original random forest algorithm and other traditional machine learning algorithms. [ABSTRACT FROM AUTHOR]

Purpose: This study aims to explore the spatio-temporal variation of soil nutrients as well as soil organic matter (SOM), and clarify the role of environmental and soil management factors in determining soil nutrients and SOM in farmland over Jiangxi Province of Southern China. Materials and methods: Between 2005 and 2012, we collected 16,504 surface soil samples (0–20 cm) from farmland across Jiangxi Province. Based on this soil dataset, we summarized the changes in SOM, alkali-hydrolyzable nitrogen (available N), available phosphorus (P), available potassium (K), pH, and cation exchange capacity. Then, we used the geostatistical method to explore and map the spatio-temporal variability of SOM, and available N, P, and K. Finally, the random forest algorithm was used to identify the main factors controlling the variation of SOM and available N, P, and K. Results and discussion: Our results revealed a clear right-skewed trend for the histogram of available P and K, pH, and cation exchange capacity in farmland soil of Jiangxi Province. From 2005 to 2012, the average concentrations of SOM and available P showed an insignificant decreasing temporal trend in the farmland of Jiangxi Province. The average concentrations of available N and available K showed a significant increasing trend between 2005 and 2012. In general, most of the soil samples had SOM, available N and P content at or above the level of Class 3 (high grade), and available K at or below the level of class 4 (moderate grade). The apparent lack of K fertility was detected. Regarding the spatio-temporal variation pattern, noticeable changes in the concentrations of SOM and available N, P, and K were detected in most of the region in 2012 when compared with 2005. Conclusion: Farmland soils in Jiangxi Province had good fertility, and soil nutrients and SOM in farmland showed strong spatial variability. Overall, the climate (e.g., mean annual precipitation and mean annual temperature) and soil management (e.g., straw return and chemical fertilizer application) had dominant effects on soil nutrients and SOM, while other factors such as relief and soil properties had slight effect. The straw return is a sustainable way to improve soil fertility. Moreover, soil pH has a slight impact on soil nutrients and SOM. Great efforts are needed to prevent farmland soils from further acidification. [ABSTRACT FROM AUTHOR]

The glass relics are precious material evidence of the early trade and cultural exchange between the East and the West. To explore the cultural differences and trade development between early China and foreign countries, it is extremely important to classify glass cultural relics. Despite their similar appearances, Chinese glass contains more lead, while foreign glass contains more potassium. In view of this, this paper proposes a joint Daen-LR, ARIMA-LSTM, and MLR machine learning algorithm (JMLA) for the analysis and identification of the chemical composition of ancient glass. We separate the sampling points of ancient glass into two systems: lead-barium glass and high-potassium glass. Firstly, an improved logistic regression model based on a double adaptive elastic network (Daen-LR) is used to select variables with both Oracle and adaptive classification characteristics. Secondly, the ARIMA-LSTM model was used to establish the correlation curve of chemical composition before and after weathering and to predict the change in chemical composition with weathering. Thirdly, combining the data processed by the above two methods, a multiple linear regression model (MLR) is used to classify unknown glass products. It was shown that the sample obtained by this processing method has a very good fit. In comparison with other similar types of models like Decision Trees (DT), Random Forests (RF), Support Vector Machines (SVM), and Random Forests based on classification and regression trees (CART-RF), the classification accuracy of JMLA is 97.9% on the train set. The accuracy rate on the test set reached 97.6%. The results of the research demonstrate that JMLA can improve the accuracy of the glass type classification problem, greatly enhance the research efficiency of archaeological staff, and gain a more reliable result. [ABSTRACT FROM AUTHOR]

In the mapping and assessment of mountain hazard susceptibility using machine learning models, the selection of model parameters plays a critical role in the accuracy of predicting models. In this study, we present a novel approach for developing a prediction model based on random forest (RF) by incorporating ensembles of hyperparameter optimization. The performance of the RF model is enhanced by employing a Bayesian optimization (Bayes) method and a genetic algorithm (GA) and verified in the Wudu section of the Bailong River basin, China, which is a typical hazard-prone, mountainous area. We identified fourteen influential factors based on field measurements to describe the "avalanche–landslide–debris flow" hazard chains in the study area. We constructed training (80%) and validation (20%) datasets for 378 hazard sites. The performance of the models was assessed using standard statistical metrics, including recall, confusion matrix, accuracy, F1, precision, and area under the operating characteristic curve (AUC), based on a multicollinearity analysis and Relief-F two-step evaluation. The results indicate that all three models, i.e., RF, GA-RF, and Bayes-RF, achieved good performance (AUC: 0.89~0.92). The Bayes-RF model outperformed the other two models (AUC = 0.92). Therefore, this model is highly accurate and robust for mountain hazard susceptibility assessment and is useful for the study area as well as other regions. Additionally, stakeholders can use the susceptibility map produced to guide mountain hazard prevention and control measures in the region. [ABSTRACT FROM AUTHOR]

Accurate information on dominant tree species and their spatial distribution in subtropical natural forests are key ecological monitoring factors for accurately characterizing forest biodiversity, depicting the tree competition mechanism and quantitatively evaluating forest ecosystem stability. In this study, the subtropical natural forest in northwest Yunnan province of China was selected as the study area. Firstly, an object-oriented multi-resolution segmentation (MRS) algorithm was used to segment individual tree crowns from the UAV RGB imagery and satellite multispectral imagery in the forests with different densities (low (547 n/ha), middle (753 n/ha) and high (1040 n/ha)), and parameters of the MRS algorithm were tested and optimized for accurately extracting the tree crown and position information of the individual tree. Secondly, the texture metrics of the UAV RGB imagery and the spectral metrics of the satellite multispectral imagery within the individual tree crown were extracted, and the random forest algorithm and three deep learning networks constructed in this study were utilized to classify the five dominant tree species. Finally, we compared and evaluated the performance of the random forest algorithm and three deep learning networks for dominant tree species classification using the field measurement data, and the influence of the number of training samples on the accuracy of dominant tree species classification using deep learning networks was investigated. The results showed that: (1) Stand density had little influence on individual tree segmentation using the object-oriented MRS algorithm. In the forests with different stand densities, the F1 score of individual tree segmentation based on satellite multispectral imagery was 71.3–74.7%, and that based on UAV high-resolution RGB imagery was 75.4–79.2%. (2) The overall accuracy of dominant tree species classification using the light-weight network MobileNetV2 (OA = 71.11–82.22%), residual network ResNet34 (OA = 78.89–91.11%) and dense network DenseNet121 (OA = 81.11–94.44%) was higher than that of the random forest algorithm (OA = 60.00–64.44%), among which DenseNet121 had the highest overall accuracy. Texture metrics improved the overall accuracy of dominant tree species classification. (3) For the three deep learning networks, the changes in overall accuracy of dominant tree species classification influenced by the number of training samples were 2.69–4.28%. [ABSTRACT FROM AUTHOR]

The objective of the study is to explore the factors that influence the job satisfaction and organizational commitment of primary care providers in China, with a focus on the impact of the COVID-19 pandemic and the rescission of restriction policies. We utilized the 20-item Minnesota Satisfaction Questionnaire (MSQ) and the 25-item organizational commitment survey to assess job satisfaction and organizational commitment. In total, 435 valid responses were included in our analysis. The average scores for job satisfaction and organizational commitment were 80.6 and 90.8. After a two-step tuning process, we built random forest models by machine learning. The results show income change, working years, working years in the current institute, and age were the four most important features associated with job satisfaction, organizational commitment, and most of their dimensions. The number of professional fields engaged, gender, job status, and types of endowment insurance were least associated. During pandemic time, income-related factors remain a core concern for primary care providers, whereas job security may lose its importance. These findings suggest that financial bonuses may be an effective way to boost morale, and age-specific motivation plans may be necessary. [ABSTRACT FROM AUTHOR]

Intelligent compaction (IC) technology have been used for quality control and quality assurance (QC/QA) in subgrade construction. The effective regression correlations between Intelligent Compaction Measurement Values (ICMV) and In-situ Measurement Values (ISMV, including compaction degree and subgrade modulus ELWD) are the essential prerequisite of using IC technology for QC/QA. This paper presents the results from an experimental research study that was conducted from a practical subgrade project of China to explore the regression relationships between ICMV and ISMV. Three types of ICMV, including CMV, CCV and Evib, were collected along with the corresponding positions of the rollers. Two types of ISMV, containing compaction degree and ELWD, were measured by ring sampler method and light weight deflectometer (LWD) at specified test points, respectively. Based on these data, the influences of roller parameters and subgrade properties on the regression relationships of ICMV and ISMV were investigated. In addition, linear regression and 5 nonlinear regression algorithms were compared. The results suggest that ICMV reflect the stiffness of subgrade more than reflecting the density. In the regression of ICMV and ISMV, subgrade properties are more important than roller parameters while both of them should not be neglected. The influences of underlying stiffness and roller amplitude are linear while that of roller speed and moisture content are nonlinear. Nonlinear algorithms, especially the random forest, have the better performance compared to linear algorithm. Moreover, the combination of random forest and linear algorithm can further improve the accuracy of ISMV prediction. [ABSTRACT FROM AUTHOR]

The quantification of grassland above-ground biomass (AGB) and grazing intensity (GI) and their distribution in space is of great significance to grassland management and eco-conservation. Remote-sensing technology is widely applied, but it is difficult to measure accurately when monitoring GI. In this study, the neural network, random forest and statistical function models of the relationship between Landsat NDVI and AGB were constructed by field survey and literature data collection in Inner Mongolia grassland, China. By comparing the accuracy among the three models, we constructed a remote-sensing retrieving model of grass AGB. We also estimated the grassland AGB during the peak growing season (August) for Inner Mongolia. Frequency histograms were then made to identify AGB thresholds under four GI levels (light or ungrazed, moderate grazing, overgrazing and extreme grazing) for each of three grassland types (meadow steppe, typical steppe and desert steppe). This study shows that the random forest model simulates grass AGB more accurately than other models. The spatial distribution of AGB in Inner Mongolia grasslands showed a tendency of decreasing from southeast to northwest, with an increasing trend in the last 10 years. The four GI levels in 2021 accounted for 18%, 25%, 36% and 21% of the grasslands in Inner Mongolia, respectively, and over the last 10 years the GI first improved and then deteriorated. This study provides a guideline to remote monitoring for grassland AGB and GI, and supplies scientific support for sustainable management and grassland restoration of large-scale grasslands. [ABSTRACT FROM AUTHOR]

Agricultural drought significantly impacts agricultural production, highlighting the need for accurate monitoring. Accurate agricultural-drought-monitoring models are critical for timely warning and prevention. The random forest (RF) is a popular artificial intelligence method but has not been extensively studied for agricultural drought monitoring. Here, multi-source remote sensing data, including surface temperature, vegetation index, and soil moisture data, were used as independent variables; the 3-month standardized precipitation evapotranspiration index (SPEI_3) was used as the dependent variable. Soil texture and terrain data were used as auxiliary variables. The bias-corrected RF model was used to construct a random forest synthesized drought index (RFSDI). The drought-degree determination coefficients (R2) of the training and test sets reached 0.86 and 0.89, respectively. The RFSDI and SPEI_3 fit closely, with a correlation coefficient (R) above 0.92. The RFSDI accurately reflected typical drought years and effectively monitored agricultural drought in Northeast China (NEC). In the past 18 years, agricultural drought in NEC has generally decelerated. The degree and scope of drought impacts from 2003 to 2010 were greater than those from 2010 to 2020. Agricultural drought occurrence in NEC was associated with dominant climatic variables such as precipitation (PRE), surface temperature (Ts), relative humidity (RHU), and sunshine duration (SSD), alongside elevation and soil texture differences. The agricultural drought occurrence percentage at 50–500 m elevations reached 94.91%, and the percentage of occurrence in loam and sandy soils reached 90.31%. Water and temperature changes were significantly correlated with the occurrence of agricultural drought. Additionally, NEC showed an alternating cycle of drought and waterlogging of about 10 years. These results have significant application potential for agricultural drought monitoring and drought prevention in NEC and demonstrate a new approach to comprehensively evaluating agricultural drought. [ABSTRACT FROM AUTHOR]

Over the past two decades, the location and morphology of the coastline, as well as the land use/land cover (LULC) in the Circum-Bohai region in China, have undergone significant changes due to rapid industrialization and urbanization. Analyzing the temporal and spatial variation in coastal lines and LULC can provide a meaningful basis for the rational allocation of land resources. Using Landsat TM/OLI series dates from the Google Earth Engine (GEE) platform, this study applied the Linear Superposition Water Index (LSWI) and the Otsu threshold method (OTSU) algorithm to extract and analyze the coastline of the Circum-Bohai region. Additionally, the Random Forests (RF) method was employed to extract LULC information in the coastal zone. Using the geographical detector, we further explored the influence of social and economic factors, as well as natural factors, on spatial differentiation mechanisms of LULC change in the Circum-Bohai. Our results show that between 2000 and 2020, the Circum-Bohai coastline generally expanded towards the ocean by a total of 1062.99 km. The highest rate of change occurred during 2010 to 2015, and human activities were the primary cause of most of the changes, with the exception of the Yellow River Delta, where natural factors were dominant. The main types of LULC in the study area from 2000 to 2020 were farmland and construction land. The area of farmland proportion decreased by 1.75%, while the area of construction land proportion increased from 16.73% to 29.54%. Our findings indicate that the degree of land use in the Circum-Bohai is deepening. Based on our factor detection analysis, the added value of the secondary industry was the most critical influencing factor on LULC. Furthermore, the combined effect of the added value of the secondary industry and gross domestic product (GDP) has a significant driving impact on LULC. These findings can provide reference and data support for the sustainable development and comprehensive management of land resources. The relevant departments can use these results to prompt corresponding policies for the rational allocation of land resources. [ABSTRACT FROM AUTHOR]

Runoff prediction with high accuracy is vital to protect people's properties and lives. In this study, a Short-Term Streamflow Forecasting Framework combining Long Short-Term Memory (LSTM) with Self-Organizing Map (SOM) is proposed, which clusters hydro-meteorological factors through SOM network to cluster the streamflow into several flow modes with different physical characteristics, converting complex global modeling into local modeling. The framework also considers the temporal characteristics of the streamflow, so the data of different flow modes are fitted with multiple LSTM separately. Besides, Random Forest is used in the framework for feature selection. The model is used for streamflow forecasting 1 to 3 day(s) ahead in the Qingxi River Basin in China, and the performance is compared with those of SVR, ANN, LSTM, SOM-SVR, SOM-ANN, FCM-LSTM, and K-means-LSTM. In addition, SOM-LSTM with three and four clusters were compared. The results show that (1) the sub-processes obtained by SOM clustering show some physical characteristics consistent with the actual process. (2) SOM-LSTM significantly improves the prediction accuracy of short-term streamflow forecasting. (3) All the integrated models using different clustering methods can improve the prediction accuracy of a single LSTM, and SOM-LSTM has the most reliable prediction when the lead time is 1–2 day(s). (4) The number of clusters being set to 4 is more suitable for the proposed prediction framework. The overall results show that the SOM-LSTM provides an efficient method for short-term streamflow forecasting. [ABSTRACT FROM AUTHOR]

Accurate information on crop type spatial distributions is crucial for yield estimation, agroecological modeling, and agrarian policy development. In comparison with moderate resolution (10–30 m) satellite images, very high-resolution (VHR, ≤ 1 m) images are superior at describing the spatial details and complex texture features of agricultural land. However, VHR images generally suffer from insufficient spatiotemporal coverage and high spectral variation in crop type, making fine-scale crop type mapping at regional or larger scales challenging. In this study, we developed a novel data-driven residual attention U-shape network (RAUNet) to identify crop types at the regional scale based on VHR images. RAUNet was adapted from UNet by introducing a residual module and a gated attention mechanism to strengthen the multilevel representation of crop features from VHR images. We selected rice-crayfish fields (RCFs) in Hubei Province, China, as the case crop type and area to test the performance of RAUNet. Seven models, including RAUNet pan , ResUNet, AttnUNet, UNet, DeepLabV3 + and random forest at the pixel and object levels, were adopted for comparison with RAUNet. Moreover, the model transferability was evaluated over regions with different agricultural landscapes. Our results showed that RAUNet-derived RCF maps obtained an average F1-score and MCC of 0.90 and 0.85 at different cultivation stages, which significantly outperformed other methods. Additionally, RAUNet had good spatiotemporal transferability, with an average F1-score and MCC of 0.93 and 0.88 in four transfer areas. The visualization results of the attention weights indicated RAUNet's prominent capability in combining useful and relevant information from different feature levels to produce refinement predictions for the target crop type. Furthermore, the high-level features derived from RAUNet can well separate RCFs and non-RCFs over different landscapes, which were also consistent across different cultivation periods. These encouraging results suggest that RAUNet works well for fine-scale crop type mapping in smallholder farming systems, and it is an efficient method for land cover mapping based on VHR images at the regional scale. [ABSTRACT FROM AUTHOR]

Hyperspectral remote sensing technology has been widely used in vegetation type mapping, and the integrated use of spectral and texture features of remote sensing data has become an effective way to improve the accuracy of vegetation classification, but the Hughes effect generated by the high-dimensional feature space composed of multi-scale texture features and hyperspectral features reduced the image classification accuracy, and the randomness of feature optimization algorithms such as genetic algorithm (GA) can also cause uncertainty in image classification accuracy. In order to solve those problems, a high-dimensional spectral-texture feature space was established by incorporating spectral features of ZY1 02D Advanced HyperSpectral Imager(AHSI) hyperspectral data and multi-scale texture features extracted from ZY1 02D Visible and Near Infrared Camera(VNIC) data at various window sizes and movement directions, covering the vicinity of Zongjia Town in Dulan County, Qinghai Province, China. Jeffries-Matusita(J-M) distance is employed to assess the sample separability between vegetation types of texture features derived from varying window sizes and movement directions. Based on a combination of feature optimization algorithms (GA algorithm, particle swarm optimization (PSO), generalized normal distribution optimization (GNDO), atom search algorithm (ASO) and marine predators algorithm (MPA)) and classification algorithm (random forest (RF)), GA-RF, PSO-RF, GNDO-RF, ASO-RF, and MPA-RF algorithms were proposed and applied to the vegetation type classification of the high-dimensional spectral-texture feature space. The results show that the texture features extracted from different window sizes and window movement directions showed maximum J-M distance between distinct vegetations. For instance, the J-M distance between wolfberry (old) and wolfberry (new) is highest in the 3×3 window size texture images, while the J-M distance between wolfberry (old) and haloxylon is highest in the 11×11 window size texture images. In the 0° window movement direction texture images, poplar and haloxylon had the best separability to other vegetation types with J-M distances greater than 1.87 and 1.86, respectively. Grass and other vegetation types achieved the best separability with J-M distances greater than 1.18 in the 90° window movement direction texture images. The inclusion of multiscale texture features improved the overall classification accuracy (OA) by 8.02 percentage. The proposed GA-RF, PSO-RF, GNDO-RF, ASO-RF, and MPA-RF algorithms, when compared to the conventional random forest algorithm, led to an improvement in OA of vegetation classification by 1.32 to 2.41 percentage. Among them, the MPA-RF method exhibited the highest accuracy, achieving an OA and Kappa coefficient of 88.92% and 0.86, respectively. The results of this study demonstrate that texture images extracted from different window sizes and window movement directions are useful in distinguishing between different types of vegetation. The accuracy of vegetation recognition is significantly increased by adding multi-scale texture features to the spectral feature set. The combination of feature optimization algorithm and image classification algorithm by iterative optimization alleviates the randomness of the optimization algorithm results, overcomes the Hughes effect of high-dimensional features, and successfully improves the vegetation classification accuracy. This study proposes innovative ideas for feature extraction, feature optimization, and classification algorithm selection for hyperspectral vegetation classification studies. [ABSTRACT FROM AUTHOR]

In recent years, the development of machine learning, especially deep learning, has provided new methods and ideas for current climate research. Stacked auto-encoder based on deep learning are used to perform nonlinear down-scaling to compress the degree of freedom of climate variables in the early stage. And climate predictors features are extracted from the summer precipitation in four regions in eastern China, from which key climate predictors affecting summer precipitation in each region are identified. On this basis, a variety of regression methods including random forest in machine learning method were used to construct prediction models for key climate predictors in each region. The best model parameters were determined by the sensitivity test of model parameters to forecast results. Several years of predictions suggest that the method for the forecast of precipitation in eastern China has a very high skill, especially in southern China. The results show that the anomaly consistency of the proposed model in regional prediction is better than that of the model. Compared with the mainstream model, the prediction results in South China can be improved by more than 10%. The method has a good application prospect for summer precipitation prediction in eastern China. [ABSTRACT FROM AUTHOR]

Landslide susceptibility mapping of mountain roads is frequently confronted by insufficient historical landslide sample data, multicollinearity of existing evaluation index factors, and inconsistency of evaluation factors due to regional environmental variations. Then, a single machine learning model can easily become overfitting, thus reducing the accuracy and robustness of the evaluation model. This paper proposes a combined machine-learning model to address the issues. The landslide susceptibility in mountain roads were mapped by using factor analysis to normalize and reduce the dimensionality of the initial condition factor and generating six new combination factors as evaluation indexes. The mountain roads in the Youxi County, Fujian Province, China were used for the landslide susceptibility mapping. Three most frequently used machine learning techniques, support vector machine (SVM), random forest (RF), and artificial neural network (ANN) models, were used to model the landslide susceptibility of the study area and validate the accuracy of this evaluation index system. The global minimum variance portfolio was utilized to construct a machine learning combined model. 5-fold cross-validation, statistical indexes, and AUC (Area Under Curve) values were implemented to evaluate the predictive accuracy of the landslide susceptibility model. The mean AUC values for the SVM, RF, and ANN models in the training stage were 89.2%, 88.5%, and 87.9%, respectively, and 78.0%, 73.7%, and 76.7%, respectively, in the validating stage. In the training and validation stages, the mean AUC values of the combined model were 92.4% and 87.1%, respectively. The combined model provides greater prediction accuracy and model robustness than one single model. [ABSTRACT FROM AUTHOR]

This study uses five machine learning algorithms (Stochastic gradient descent (SGD), Decision tree, Random forest, Gradient boosting decision tree (GBDT), and Convolutional neural networks (CNN)) to explore their prediction effects on China's stock market. It constructs a monthly rolling model for stock return prediction. Selecting stocks of the CSI 300 index from January to June 2021 as specific monthly samples and classifying three factors – fundamentals, volatility(risk) and technical indicators, the results demonstrate that (1) machine learning brings favorable investment returns in simulated quantitative trading of China's A-share market (2) the technical indicator factor is the most valuable, with the momentum technical factor having greatest influence, followed by the volatility (risk) factor and fundamental factors. Therefore, this study has a critical reference value and is significant in guiding yield forecasting in intricate stock markets. [ABSTRACT FROM AUTHOR]

Purpose: The purpose of this paper is to detect predefined service attributes and their sentiments from online restaurant reviews, and then to measure the effects of customer sentiments toward service attributes on customer satisfaction (CS) and revisit intention (RVI) simultaneously. Design/methodology/approach: This study proposed a supervised framework to model CS and RVI simultaneously from restaurant reviews. Specifically, the authors detected the predefined service dimensions from online reviews based on random forest. Then, the sentiment polarities of the reviews toward each predefined dimension were identified using light-gradient boosting machine (LightGBM). Finally, the effects of attribute-specific sentiments on CS and RVI were evaluated by a bagged neural network-based model. The proposed framework was evaluated by 305,000 restaurant comments collected from DianPing.com, a Yelp-like website in China. Findings: The authors obtained a hierarchal importance order of the investigated service themes (i.e. location, service, environment, price and food). The authors found that food played the most important role in affecting both CS and RVI. The most salient attribute with respect to each service theme was also identified. Originality/value: Unlike prior work relying on the data collected from surveys, this study is among the first to model the relationship among service attributes, CS and RVI simultaneously from real-world data. The authors established a hierarchal structure of eighteen attributes within five service themes and estimated their effects on both CS and RVI, which will broaden our understanding of customer perception and behavioral intention during service consumption. [ABSTRACT FROM AUTHOR]

To effectively explore the impacts of several key factors on taxi drivers' traffic violations and provide traffic management departments with scientific decisions to reduce traffic fatalities and injuries. 43,458 electronic enforcement data about taxi drivers' traffic violations in Nanchang City, Jiangxi Province, China, from July 1, 2020, to June 30, 2021, were utilized to explore the characteristics of traffic violations. A random forest algorithm was used to predict the severity of taxi drivers' traffic violations and 11 factors affecting traffic violations, including time, road conditions, environment, and taxi companies were analyzed using the Shapley Additionality Explanation (SHAP) framework. Firstly, the ensemble method Balanced Bagging Classifier (BBC) was applied to balance the dataset. The results showed that the imbalance ratio (IR) of the original imbalanced dataset reduced from 6.61% to 2.60%. Moreover, a prediction model for the severity of taxi drivers' traffic violations was established by using the Random Forest, and the results showed that accuracy, m_F1, m_G-mean, m_AUC, and m_AP obtained 0.877, 0.849, 0.599, 0.976, and 0.957, respectively. Compared with the algorithms of Decision Tree, XG Boost, Ada Boost, and Neural Network, the performance measures of the prediction model based on Random Forest were the best. Finally, the SHAP framework was used to improve the interpretability of the model and identify important factors affecting taxi drivers' traffic violations. The results showed that functional districts, location of the violation, and road grade were found to have a high impact on the probability of traffic violations; their mean SHAP values were 0.39, 0.36, and 0.26, respectively. Findings of this paper may help to discover the relationship between the influencing factors and the severity of traffic violations, and provide a theoretical basis for reducing the traffic violations of taxi drivers and improving the road safety management. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Irrigation & Drainage is the property of Journal of Irrigation & Drainage Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Wildfires directly threaten the safety of life and property. Predicting wildfires with a model driven by wildfire danger factors can significantly reduce losses. Weather conditions continuously influence the drying rate of fuel as well as the occurrence probability and danger degree of wildfires. Previous studies have paid little attention to the continuous effects of weather and fuel on wildfires. This study improved the accuracy and effect of wildfire danger assessment using the time series features of weather and fuel. First, the time series features of weather and fuel factors within the 16 days before the fire were analyzed. Then, four feature groups were selected—feature group without time series values, feature group with time series values, feature group with Tsfresh transformation of time series values, and feature group with gradient and cumulative transformation of time series values—and three models were trained, respectively: random forest, balanced random forest, and extreme gradient boosting. The results showed that the f1-score of all feature groups with time series values (0.93) increased by 0.15, on average, compared with those without time series values (0.78) for the three models. The feature group with gradient and cumulative features had a more stable prediction accuracy and a more accurate wildfire danger map. The results suggest that using the appropriate time series features of weather and fuel can help improve the precision and effect of the wildfire danger assessment model. [ABSTRACT FROM AUTHOR]

Timely wildfire detection is helpful for fire monitoring and fighting. However, the available wildfire products with high temporal resolutions face problems, such as high omission error and commission error (false alarm) rates. This study proposed a wildfire detection algorithm combined with an improved spatial contextual algorithm and machine learning method in southwest China. First, a dataset consisting of a formation of high-confidence fire pixels combining the WLF (Himawari Wild Fire product) and VIIRS wildfire products was constructed. Then, a model to extract potential fire pixels was built using the random forest method. Additionally, an improved spatial contextual algorithm was used to identify actual fire pixels from potential fire pixels. Finally, strategies such as sun glint rejection were used to remove false alarms. As a result, the proposed algorithm performed better, with both a lower omission error rate and a lower commission error rate than the WLF product. It had a higher F1 score (0.47) than WLF (0.43) with VIIRS for reference, which means it is more suitable for wildfire detection. [ABSTRACT FROM AUTHOR]

With the development of the Internet of Things (IoT) and virtual reality (VR) technology, the demand for high-precision gesture intelligent analysis of a human–machine interaction module for IoT–VR systems is increasing. Therefore, random forest (RF) and convolution neural network (CNN) algorithms are used in this study to build an intelligent gesture recognition model. The experiments were conducted to test the application performance of the design model. The test results show that the qualification rate of the analytical model designed in this study is significantly higher than that of the comparative model. When the threshold is determined to be 43.26 mm, the analytical qualification rates of the RF-CNN (the method of combining RF with CNN algorithms), faster regions with CNN features (Faster-RCNN), and RF models are 82.41%, 76.10%, and 59.10%, respectively. The calculation time of the RF–CNN model is between the two comparative models. From the test data, it can be observed that the research results have certain significance for improving the accuracy of gesture machine recognition technology in China's VR Internet of Things (IoT) system. [ABSTRACT FROM AUTHOR]

Frost damage to winter wheat during stem elongation frequently occurred in the Huang-Huai plain of China, leading to considerable yield losses. Minimum Stevenson screen temperature (STmin) and minimum grass temperature (GTmin) have long been used to quantify frost damage. Although GTmin has higher accuracy than STmin, it is limited in application due to the lack of data. Therefore, this study aimed to select appropriate environmental variables to estimate GTmin, as well as to quantify the frost damage. Shangqiu, a frost-prone winter wheat area in the central Huang-Hui plain, was selected as the study area. From the descriptive statistics of ST, air relative humidity (RH), wind speed (WS), cloud fraction (CF), and volumetric soil water content (VWC) during temperature decreasing and increasing, seven variables significantly correlated with GTmin were selected, including STmin, maximum reduction of ST (RST), maximum increase of ST (IST), minimum RH during temperature increasing (RHmin), WS at STmin occurrence (WS), minimum VWC during temperature decreasing (VWCmin), and nightly CF. Multiple linear regression (MLR), support vector regression (SVR), random forest (RF), and K-nearest neighbor (KNN) were adopted for estimating GTmin based on the various combinations of the variables. Results showed the more variables, the higher the accuracy for the MLR and SVR. However, this pattern was not always true for the KNN and RF. The KNN based on STmin, RST, IST, RHmin, and WS achieved the highest accuracy, with R2 of 0.9992, RMSE of 0.14 ℃, and MAE of 0.076 ℃. The overall classification accuracy for frost damage identified by the estimated GTmin reached 97.1% during stem elongation of winter wheat from 2017 to 2021. The integrated frost stress (IFS) index calculated by the estimated and measured GTmin maintained high linear fitting accuracy. The KNN with fewer variables demonstrated good applicability at the regional scale. [ABSTRACT FROM AUTHOR]

Using the data of Chinese listed enterprises in 2007–2020, we investigate the relationship between green innovation (GI) and enterprise reputation via OLS regression, IV‐Lasso, and random forest. Findings show that GI can improve enterprise reputation, and it has a significant positive lag effect. Additionally, we compare the different effects of green patent applications and approved green patents. A possible mechanism is that GI reduces the pollution emission of enterprises. GI also performs better in improving the reputation of coastal enterprises, state‐owned enterprises and large‐scale enterprises, and old enterprises. Our finding on the causal relationship between GI and enterprise reputation is of great significance in further understanding the practical value of GI activities, and it has certain reference value for the policy formulation of governments and decision making of enterprises. [ABSTRACT FROM AUTHOR]

Objective: To investigate the relationship between physical activity and the severity of menopausal symptoms in middle-aged women in northwest China. Methods: This was a cross-sectional online survey study. Using a snowball sampling method, 468 women aged 45 to 60 were recruited from northwest China and their demographic information was collected. The modified Kupperman Menopausal Index scale and International Physical Activity Questionnaire short form were used in this study. Random forest was used to rank the importance of variables and select the optimal combination. The direction and relative risk (odds ratio value) of selected variables were further explained with an ordinal logistic regression model. Results: The prevalence of menopausal syndromes was 74.8% and more than one-half of the participants had moderate or severe symptoms (54.3%). The Mantel-Haenszel linear-by-linear chi-square test showed a strong and negative correlation between physical activity level and the severity of menopausal symptoms (P < 0.001). Random forest demonstrated that the physical activity level was the most significant variable associated with the severity of menopausal symptoms. Multiple random forest regressions showed that the out-of-bag error rate reaches the minimum when the top 4 variables (physical activity level, menopausal status, perceived health status, and parity) in the importance ranking form an optimal variable combination. Ordinal logistic regression analysis showed that a higher physical activity level and a satisfactory perceived health status might be protective factors for menopausal symptoms (odds ratio (OR) < 1, P < 0.001); whereas perimenopausal or postmenopausal status and 2 parities might be risk factors for menopausal symptoms (OR > 1, P < 0.001). Conclusions: There is a strong negative correlation between physical activity and the severity of menopausal symptoms. The results have a clinical implication that the menopausal symptoms may be improved by the moderate-to-high level physical activity in the lives of middle-aged women. [ABSTRACT FROM AUTHOR]

Landslides occur in most countries. As one of the most serious geological hazards, landslides affect infrastructure construction. Thus, it is vital to prepare reliable landslide susceptibility evaluation maps to avoid landslide-prone areas for various construction projects. In recent years, supervised machine learning algorithms have been widely used in landslide susceptibility evaluation, but many flaws remain in the selection of nonlandslide point samples for comparative analysis. It is significant to improve the authenticity of sample data and reduce the impact of noise. China's Funing County was used as a case study in this paper, which first identified 122 landslide incidents based on historical data, fieldwork, and remote sensing images to create a landslide inventory in the research area. In addition, 12 causal factors of landslides were determined, including elevation, slope, aspect, plan curvature, profile curvature, distance to roads, distance to rivers, distance to faults, rainfall, normalized difference vegetation index (NDVI), lithology, and land cover. K-means clustering was used to purify the factor data factors before data-driven certainty factor (CF) and frequency ratio (FR) models, and machine learning models, random forest (RF) and artificial neural network (ANN), were used for a comparative study on landslide susceptibility evaluation in Funing County. The results show that the selection method of nonlandslide sample data will affect the accuracy of different evaluation models. The purified sample data improved the prediction accuracy of the four models, with significant prediction accuracy improvements observed in the ANN model. The purification of nonlandslide sample data by K-means method is of great significance for the drawing of landslide sensitivity map. [ABSTRACT FROM AUTHOR]

Crop residue cover is vital for reducing soil erosion and improving soil fertility, which is an important way of conserving tillage to protect the black soil in Northeast China. How much the crop residue covers on cropland is of significance for black soil protection. Landsat-8 and Sentinel-2 images were used to estimate corn residue coverage (CRC) in Northeast China in this study. The estimation model of CRC was established for improving CRC estimation accuracy by the optimal combination of spectral indices and textural features, based on soil texture zoning, using the random forest regression method. Our results revealed that (1) the optimization C5 of spectral indices and textural features improves the CRC estimation accuracy after harvesting and before sowing with determination coefficients (R2) of 0.78 and 0.73, respectively; (2) the random forest improves the CRC estimation accuracy after harvesting and before sowing with R2 of 0.81 and 0.77, respectively; (3) considering the spatial heterogeneity of the soil background and the usage of soil texture zoning models increase the accuracy of CRC estimation after harvesting and before sowing with R2 of 0.84 and 0.81, respectively. In general, the CRC estimation accuracy after harvesting was better than that before sowing. The results revealed that the corn residue coverage in most of the study area was 0.3 to 0.6 and was mainly distributed in the Songnen Plain. By the estimated corn residue coverage results, the implementation of conservation tillage practices is identified, which is vital for protecting the black soil in Northeast China. [ABSTRACT FROM AUTHOR]

Accurately estimating land-use demand is essential for urban models to predict the evolution of urban spatial morphology. Due to the uncertainties inherent in socioeconomic development, the accurate forecasting of urban land-use demand remains a daunting challenge. The present study proposes a modeling framework to determine the scaling relationship between the population and urban area and simulates the spatiotemporal dynamics of land use and land cover (LULC). An allometric scaling (AS) law and a Markov (MK) chain are used to predict variations in LULC. Random forest (RF) and cellular automata (CA) serve to calibrate the transition rules of change in LULC and realize its micro-spatial allocation (MKCARF-AS). Furthermore, this research uses several shared socioeconomic pathways (SSPs) as scenario storylines. The MKCARF-AS model is used to predict changes in LULC under various SSP scenarios in Jinjiang City, China, from 2020 to 2065. The results show that the figure of merit (FoM) and the urban FoM of the MKCARF-AS model improve by 3.72% and 4.06%, respectively, compared with the MKCAANN model during the 2005–2010 simulation period. For a 6.28% discrepancy between the predicted urban land-use demand and the actual urban land-use demand over the period 2005–2010, the urban FoM degrades by 21.42%. The growth of the permanent urban population and urban area in Jinjiang City follows an allometric scaling law with an exponent of 0.933 for the period 2005–2020, and the relative residual and R2 are 0.0076 and 0.9994, respectively. From 2020 to 2065, the urban land demand estimated by the Markov model is 19.4% greater than the urban area predicted under scenario SSP5. At the township scale, the different SSP scenarios produce significantly different spatial distributions of urban expansion rates. By coupling random forest and allometric scaling, the MKCARF-AS model substantially improves the simulation of urban land use. [ABSTRACT FROM AUTHOR]

This study investigated the accuracy of the random forest (RF) model in gap filling the sensible (H) and latent heat (LE) fluxes, by using the observation data collected at a site over rice–wheat rotation croplands in Shouxian County of eastern China from 15 July 2015 to 24 April 2019. Firstly, the variable significance of the machine learning (ML) model's five input variables, including the net radiation (Rn), wind speed (WS), temperature (T), relative humidity (RH), and air pressure (P), was examined, and it was found that Rn accounted for 78 % and 76 % of the total variable significance in H and LE calculating, respectively, showing that it was the most important input variable. Secondly, the RF model's accuracy with the five-variable (Rn , WS, T , RH, P) input combination was evaluated, and the results showed that the RF model could reliably gap fill the H and LE with mean absolute errors (MAEs) of 5.88 and 20.97 W m -2 , and root mean square errors (RMSEs) of 10.67 and 29.46 W m -2 , respectively. Thirdly, four-variable input combinations were tested, and it was found that the best input combination was (Rn , WS, T , P) by removing RH from the input list, and its MAE values of H and LE were reduced by 12.65 % and 7.12 %, respectively. At last, through the Taylor diagram, H and LE gap-filling accuracies of the RF model, the support vector machine (SVM) model, the k nearest-neighbor (KNN) model, and the gradient boosting decision tree (GBDT) model were intercompared, and the statistical metrics showed that RF was the most accurate for both H and LE gap filling, while the LR and KNN model performed the worst for H and LE gap filling, respectively. [ABSTRACT FROM AUTHOR]

In recent years, extreme events and risks have increased under the background of global warming, which influenced agricultural production significantly. Adaptation has been considered as a key pathway to enhancing smallholders' climate resilience. We selected a total of 903 smallholders using the multi-stage random sampling technique in rainfed areas in China, and then collected the survey data through the structured questionnaire and focus group discussion. Three resilience-building adaptation indices (RBAS-A, RBAS-D and RBAS-I) were constructed by the entropy method, whose mean values were 0.378, 0.336, and 0.602, respectively. Furtherly, the random forest model was used to explore influencing factors of climate-resilient adaptation strategies. The results revealed that education level of household head, family size, farmland size, access to information by mass media and kith and kin, perception of temperature change in summer and winter, and perception of crop yield change were the significant factors influencing smallholders' alteration strategies. Meanwhile, age and education level of household heads, off-farm income, farmland size, mass media, and perception of winter temperature changes had significant effects on the diversification strategies. Moreover, demographic characteristics, socioeconomic characteristics, information access and climate change perceptions, and impacts had significant impacts on intensification strategies. Importantly, we found that there was a certain threshold for the impact of several factors on the constructed composite indices. And the impacts of the perceptions of temperature on alteration and intensification strategies showed a V-shape. Finally, we proposed targeted suggestions for improving smallholders' climate-resilient adaptation in the rainfed agricultural areas in China. [ABSTRACT FROM AUTHOR]

It is difficult to accurately estimate the incidence rate of intimate partner violence (IPV) using traditional social survey methods because IPV victims are often reluctant to disclose their experiences, leading to an underestimation of the incidence rate. To address this issue, we applied machine learning algorithms to predict the incidence rate of IPV in China based on data from the Third Wave Survey on the Social Status of Women in China (TWSSSCW 2010). Specifically, we examined five unbalanced sample-processing methods and six machine learning algorithms, choosing the random under-sampling ensemble method and the random forest algorithm to impute the missing data. Analysis of the complete data showed that the incidence rates of physical violence, verbal violence, and cold violence were 7.10%, 13.74%, and 21.35%, respectively, which were higher than the incidence rates in the original dataset (4.05%, 11.21%, and 17.95%, respectively). The robustness of our findings was further confirmed by analysis using different training sets. Overall, this study demonstrates that better tools need to be developed to accurately estimate the incidence rates of IPV. It also serves as a useful guide for future research that imputes missing data using machine learning. [ABSTRACT FROM AUTHOR]

Due to the rapid development of industrialization and substantial economy, China has become one of the global hotspots of nitrogen (N) and sulfur (S) deposition following Europe and the USA. Here, we developed a dataset with full coverage of N and S deposition from 2005 to 2020, with multiple statistical models that combine ground-level observations, chemistry transport simulations, satellite-derived vertical columns, and meteorological and geographic variables. Based on the newly developed random forest method, the multi-year averages of dry deposition of OXN, RDN and S in China were estimated at 10.4, 14.4 and 16.7 kg N/S ha−1 yr−1, and the analogous numbers for total deposition were respectively 15.2, 20.2 and 25.9 kg N/S ha−1 yr−1 when wet deposition estimated previously with a GAM model was included. The Rdry/wet of N stabilized in earlier years and then gradually increased especially for RDN, while that of S declined for over ten years and then slightly increased. RRDN/OXN was estimated to be larger than 1 for the whole research period and clearly larger than that of the USA and Europe, with a continuous decline from 2005 to 2011 and a more prominent rebound afterwards. Compared with the USA and Europe, a more prominent lagging response of OXN and S deposition to precursor emission abatement was found in China. The OXN dry deposition presented a descending gradient from east to west, while the S dry deposition a descending gradient from north to south. After 2012, the OXN and S deposition in eastern China declined faster than the west, attributable to stricter emission controls. Positive correlation was found between regional deposition and emissions, while smaller deposition to emission ratios (D/E) existed in developed eastern China with more intensive human activities. [ABSTRACT FROM AUTHOR]

• O 3 pollution in Tianjin was largely driven by the changes in precursors emissions. • Meteorology alleviated O 3 pollution in 2015-2016 but aggravated O 3 thereafter. • Temperature is the most important meteorological factor governing O 3 pollution. Deterioration of surface ozone (O 3) pollution in Northern China over the past few years received much attention. For many cities, it is still under debate whether the trend of surface O 3 variation is driven by meteorology or the change in precursors emissions. In this work, a time series decomposition method (Seasonal-Trend decomposition procedure based on Loess (STL)) and random forest (RF) algorithm were utilized to quantify the meteorological impacts on the recorded O 3 trend and identify the key meteorological factors affecting O 3 pollution in Tianjin, the biggest coastal port city in Northern China. After "removing" the meteorological fluctuations from the observed O 3 time series, we found that variation of O 3 in Tianjin was largely driven by the changes in precursors emissions. The meteorology was unfavorable for O 3 pollution in period of 2015-2016, and turned out to be favorable during 2017-2021. Specifically, meteorology contributed 9.3 µg/m3 O 3 (13%) in 2019, together with the increase in precursors emissions, making 2019 to be the worst year of O 3 pollution since 2015. Since then, the favorable effects of meteorology on O 3 pollution tended to be weaker. Temperature was the most important factor affecting O 3 level, followed by air humidity in O 3 pollution season. In the midday of summer days, O 3 pollution frequently exceeded the standard level (>160 µg/m3) at a combined condition with relative humidity in 40%-50% and temperature > 31°C. Both the temperature and the dryness of the atmosphere need to be subtly considered for summer O 3 forecasting. [Display omitted] [ABSTRACT FROM AUTHOR]

Forest stock volume (FSV) is a major indicator of forest ecosystem health and it also plays an important part in understanding the worldwide carbon cycle. A precise comprehension of the distribution patterns and variations of FSV is crucial in the assessment of the sequestration potential of forest carbon and optimization of the management programs of the forest carbon sink. In this study, a novel vegetation index based on Sentinel-2 data for modeling FSV with the random forest (RF) algorithm in Helan Mountains, China has been developed. Among all the other variables and with a correlation coefficient of r = 0.778, the novel vegetation index (NDVIRE) developed based on the red-edge bands of the Sentinel-2 data was the most significant. Meanwhile, the model that combined bands and vegetation indices (bands + VIs-based model, BVBM) performed best in the training phase (R2 = 0.93, RMSE = 10.82 m3ha−1) and testing phase (R2 = 0.60, RMSE = 27.05 m3ha−1). Using the best training model, the FSV of the Helan Mountains was first mapped and an accuracy of 80.46% was obtained. The novel vegetation index developed based on the red-edge bands of the Sentinel-2 data and RF algorithm is thus the most effective method to assess the FSV. In addition, this method can provide a new method to estimate the FSV in other areas, especially in the management of forest carbon sequestration. [ABSTRACT FROM AUTHOR]

Flash floods are a frequent and highly destructive natural hazard in China. In order to prevent and manage these disasters, it is crucial for decision-makers to create GIS-based flash flood susceptibility maps. In this study, we present an improved Blending approach, RF-Blending (Reserve Feature Blending), which differs from the Blending approach in that it preserves the original feature dataset during meta-learner training. Our objectives were to demonstrate the performance improvement of the RF-Blending approach and to produce flash flood susceptibility maps for all catchments in Jiangxi Province using the RF-Blending approach. The Blending approach employs a double-layer structure consisting of support vector machine (SVM), K-nearest neighbor (KNN), and random forest (RF) as base learners for level-0, and the output of level-0 is utilized as the meta-feature dataset for the meta-learner in level-1, which is logistic regression (LR). RF-Blending employs the output of level-0 along with the original feature dataset for meta-learner training. To develop flood susceptibility maps, we utilized these approaches in conjunction with historical flash flood points and catchment-based factors. Our results indicate that the RF-Blending approach outperformed the other approaches. These can significantly aid catchment-based flash flood susceptibility mapping and assist managers in controlling and remediating induced damages. [ABSTRACT FROM AUTHOR]

Selenium (Se) is a beneficial microelement for humans, and its varying abundances and shortages have attracted widespread concern. The accumulation process of soil Se is quite complicated, being controlled by multiple factors. However, the influence mechanism of soil properties, climate, and topographic conditions on Se distribution is still unclear in Se-deficient areas. For this study, we collected 2804 samples from cropland soil to assess the levels of Se and the factors that influence those levels in Se-deficient areas of southwestern China. The Se content in this area (0.17 mg/kg) was less than the mean value of China as a whole (0.29 mg/kg). Moran's I index and a random forest (RF) model showed that higher Se levels were mostly observed in the southern and northern sections of the area we studied. The RF model had excellent performance in predicting soil Se content, with an accuracy of 64%. The use of Shapley additive explanations indicated that soil organic matter (SOM) and mean annual precipitation (MAP) were the critical factors determining Se distribution. The areas with high SOM and MAP showed high Se levels. The information obtained from this work can provide guidance for agricultural planning in Se-deficient areas. [ABSTRACT FROM AUTHOR]

The Ms 8.0 Wenchuan earthquake of 2008 dramatically changed the terrain surface and caused long-term increases in the scale and frequency of landslides and debris flows. The changing trend of landslides in the earthquake-affected area over the decade since the earthquake remains largely unknown. In this study, we were able to address this issue using supervised classification methods and multitemporal remote sensing images to study landslide evolution in the worst-affected area (Mianyuan River Basin) over a period of ten years. Satellite images were processed using the maximum likelihood method and random forest algorithm to automatically map landslide occurrence from 2007 to 2018. The principal findings are as follows: (1) when compared with visual image analysis, the random forest algorithm had a good average accuracy rate of 87% for landslide identification; (2) postevent landslide occurrence has generally decreased with time, but heavy monsoonal seasons have caused temporary spikes in activity; and (3) the postearthquake landslide activity in the Mianyuan River Basin can be divided into a strong activity period (2008 to 2011), medium activity period (2012 to 2016), and weak activity period (post 2017). Landslide activity remains above the prequake level, with damaging events being rare but continuing to occur. Long-term remote sensing and on-site monitoring are required to understand the evolution of landslide activity after strong earthquakes. [ABSTRACT FROM AUTHOR]

Accurate understanding of spatial distribution and variability of soil total nitrogen (TN) is critical for the site-specific nitrogen management. Based on 4337 newly obtained soil observations and 33 covariates, this study applied the random forest (RF) algorithm and modified regression kriging (RF combined with residual kriging: RFK, hereafter) model to spatially predict and map topsoil TN content in agricultural areas of Henan Province, central China. According to the RFK prediction, topsoil TN content ranged from 0.52 to 1.81 g kg−1, and the farmland with the topsoil TN contents of 1.00–1.23 g kg−1 and 0.80–1.23 g kg−1 accounted for 48.2% and 81.2% of the total farmland area, respectively. Spatially, the topsoil TN in the study area was generally higher in the west and lower in the east. By using the Boruta variable selection algorithm, soil organic matter (SOM) and available potassium contents in topsoil, nitrogen deposition, average annual precipitation, livestock discharges, and topsoil pH were identified as the main factors driving the spatial distribution and variation of soil TN in the study area. The RF and RFK models used showed the expected performance and achieved acceptable TN prediction accuracy. In comparison, RFK performed slightly better than the RF model. The R2 and RMSE achieved by the RFK model were improved by 4.5% and 4.5%, respectively, compared with that by the RF model. However, the results suggest that RFK was inferior to the RF model in quantifying prediction uncertainty and thus may have a slight disadvantage in model reliability. [ABSTRACT FROM AUTHOR]

Purpose: Despite rising incomes and reduction of extreme poverty, the feeling of being poor remains widespread. Support programs can improve well-being, but they first require identifying who are the households that judge their income is insufficient to meet their basic needs, and what factors are associated with subjective poverty. Design/methodology/approach: Households report the income level they judge is sufficient to make ends meet. Then, they are classified as being subjectively poor if their own monetary income is inferior to the level they indicated. Second, the study compares the performance of three machine learning algorithms, the random forest, support vector machines and least absolute shrinkage and selection operator (LASSO) regression, applied to a set of socioeconomic variables to predict subjective poverty status. Findings: The random forest generates 85.29% of correct predictions using a range of income and non-income predictors, closely followed by the other two techniques. For the middle-income group, the LASSO regression outperforms random forest. Subjective poverty is mostly associated with monetary income for low-income households. However, a combination of low income, low endowment (land, consumption assets) and unusual large expenditure (medical, gifts) constitutes the key predictors of feeling poor for the middle-income households. Practical implications: To reduce the feeling of poverty, policy intervention should continue to focus on increasing incomes. However, improvements in nonincome domains such as health expenditure, education and family demographics can also relieve the feeling of income inadequacy. Methodologically, better performance of either algorithm depends on the data at hand. Originality/value: For the first time, the authors show that prediction techniques are reliable to identify subjective poverty prevalence, with example from rural China. The analysis offers specific attention to the modest-income households, who may feel poor but not be identified as such by objective poverty lines, and is relevant when policy-makers seek to address the "next step" after ending extreme poverty. Prediction performance and mechanisms for three machine learning algorithms are compared. [ABSTRACT FROM AUTHOR]

Copyright of Forest Research is the property of Forest Research Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The identification of multivariate geochemical anomalies is critical in mineral exploration. Machine learning algorithms have been successfully employed in the recognition of multivariate geochemical anomalies in support of mineral exploration, owing to their strong ability to learn the complex relationship between geochemical characteristics and mineralization. However, applications of machine learning algorithms suffer from data redundancy and the curse of dimensionality. In this study, a hybrid model combining t-distributed stochastic neighbor embedding (t-SNE) and random forest (RF) was used to solve the aforementioned problems in geochemical mapping for gold exploration in the northwestern Hubei Province of China. Specifically, t-SNE was used for dimension reduction and feature extraction from the major and trace elements of geochemical survey data, and RF was used for probabilistic classification of geochemical patterns related to gold deposits. A comparative study demonstrated that the hybrid model of t-SNE + RF possesses stronger generalization ability than that of PCA + RF and pure RF. Specifically, after 15 experiments, the mean area under the receiver operator characteristic curve (AUC) values of t-SNE + RF, PCA + RF, and pure RF were 0.83, 0.65, and 0.75, respectively. These results suggest that the hybrid model combining t-SNE and RF can more efficiently recognize geochemical anomalies associated with gold mineralization. Compared with PCA, t-SNE can more effectively identify hidden information in complex and nonlinear geochemical survey data. In addition, it can reduce information redundancy and further improve the efficiency of RF for processing multidimensional geochemical survey data. The high-probability areas obtained by t-SNE + RF showed a strong spatial correlation with known gold deposits, which can provide critical clues for further prospecting in the study area. [ABSTRACT FROM AUTHOR]

Accurately identifying groundwater contamination sites is vital for groundwater protection and restoration. This study aims to use a machine learning (ML) approach to identify groundwater contamination sites with total petroleum hydrocarbons (TPH) as target contaminants in a case study of gas stations in China. Firstly, six classical ML algorithms, including logistic regression, decision tree, gradient boosting decision tree (GBDT), random forest, multi-layer perceptron, and support vector machine, were applied to develop the identification models of TPH-contaminated groundwater with 40 features and the performances were compared. The comparison results showed that the GBDT model achieves the best prediction performance, with F1 score of 1 and AUC value of 1. Next, Bayesian optimization optimized GBDT (BO-GBDT) was conducted to further decrease the training time from 19,125 s to 513 s while maintaining the same prediction performance (F1 score = 1, AUC = 1). Finally, Shapley additive explanations (SHAP) analysis was performed on the BO-GBDT model. The SHAP results displayed that the critical feature variables in the BO-GBDT model include wind, population, evaporation, total potassium in the soil, precipitation, and leakage accident. This study demonstrated that BO-GBDT is one satisfactory model to identify groundwater TPH-contamination at gas stations. The method proposed in this study has the potential to be applied to other types of groundwater contamination sites. [ABSTRACT FROM AUTHOR]

China faces tough challenges in the process of low-carbon transformation. To determine whether China can achieve its new 2030 carbon peaking and carbon intensity reduction commitments, accurate prediction of China's CO2 emissions is vital. In this paper, the random forest (RF) model was used to screen 26 carbon emission influencing factors, and seven indicators were selected as key variables for prediction. Subsequently, a three-layer back propagation (BP) neural network was constructed to forecast China's CO2 emissions and intensity from 2020 to 2040 under the 13th Five-Year Plan, 14th Five-Year Plan, energy optimization, technology breakthrough, and dual control scenarios. The results showed that energy structure factors have the most significant impact on China's CO2 emissions, followed by technology level, and economic development factors are no longer the main drivers. Under the 14th Five-Year Plan scenario, China can achieve its carbon peaking on time, reaching 10,434.082 Mt CO2 emissions in 2030. Although the new commitment to intensity reduction (over 65%) under this scenario cannot be achieved, the 14th Five-Year Plan can bring about 73.359 and 539.710 Mt of CO2 reduction in 2030 and 2040 respectively, compared to the 13th Five-Year Plan. Under the technology breakthrough and dual control scenarios, China will meet its new commitments ahead of schedule, with the dual control scenario being the optimal pathway for CO2 emissions to peak at 9860.08 Mt in 2025. It is necessary for Chinese policy makers to adjust their current strategic planning, such as accelerating the transformation of energy structure and increasing investment in R&D to achieve breakthroughs in green technologies. [ABSTRACT FROM AUTHOR]

Windows are considered as the main culprit of heat loss in buildings and the development of advanced glazing is a global necessity. Vacuum integrated photovoltaic (VPV) glazing was proven to have great air conditioning energy-saving potential, while there is a lack of real-time zero-energy potential evaluation that considers the interaction between thermal, power and daylight performances. Given this context, a daylight-electrical-thermal coupling model was developed using Python language for VPV glazing to evaluate its overall performance for optimal design towards real-time zero-energy goal. The coupling model integrated the numerical heat transfer model, electrical model, as well as decision tree and random forest algorithms for glare and illuminance prediction. All models were validated through full-scale outdoor experiments and showed acceptable errors between experimental and simulated data. Based on that, key parameters affecting the performance of the VPV glazing, namely PV coverage, emissivity of Low-E coating, pillar separation, and inner window shades were determined through sensitivity analysis. A case study of optimal design of the VPV glazing in hot summer and cold winter zone in China showed that VPV glazing with a PV coverage of 40%, a coating emissivity of 0.068, and a pillar separation of 90 mm can achieve 67% zero-energy hours throughout the year. This study can provide a validated coupled model for performance evaluation and optimization of VPV glazing for engineering applications. [ABSTRACT FROM AUTHOR]

Lithofacies identification and classification are critical for characterizing the hydrocarbon potential of unconventional resources. Although extensive applications of machine learning models in predicting lithofacies have been applied to conventional reservoir systems, the effectiveness of machine learning models in predicting clay-rich, lacustrine shale lithofacies has yet to be tackled. Here, we apply machine learning models to conventional well log data to automatically identify the shale lithofacies of Gulong Shale in the Songliao Basin. The shale lithofacies were classified into six types based on total organic carbon and mineral composition data from core analysis and geochemical logs. We compared the accuracy of Multilayer Perceptron (MLP), Support Vector Machine (SVM), Extreme Gradient Boosting (XGBoost), and Random Forest models. We mitigated the bias of imbalanced data by applying oversampling algorithms. Our results show that ensemble methods (XGBoost and Random Forest) have a better performance in shale lithofacies identification than the other models do, with accuracies of 0.868 and 0.884, respectively. The organic siliceous shale proposed to have the best hydrocarbon potential in Gulong Shale can be identified with F1 scores of 0.853 by XGBoost and 0.877 by Random Forest. Our study suggests that ensemble machine learning models can effectively identify the lithofacies of clay-rich shale from conventional well logs, providing insight into the sweet spot prediction of unconventional reservoirs. Further improvements in model performances can be achieved by adding domain knowledge and employing advanced well log data. [ABSTRACT FROM AUTHOR]

Data-driven machine learning technology can learn and extract features, a factor which is well recognized to be powerful in the warning and prediction of severe weather. With the large-scale deployment of the radar wind profile (RWP) observational network in China, dynamical variables with higher temporal and spatial resolution in the vertical become strong supports for machine-learning-based severe convection prediction. Based on the RWP mesonet that has been deployed in Beijing, this study uses the measurements from four triangles composed of six RWP stations to determine the profiles of divergence, vorticity, and vertical velocity before rainfall onsets. These dynamic feature variables, combined with cloud properties from Himawari-8 and ERA-5 reanalysis, serve as key input parameters for two rainfall forecast models based on the random forest (RF) classification algorithm. One is for the rainfall/non-rainfall forecast and another for the rainfall grade forecast. The roles of dynamic features such as divergence, vorticity, and vertical velocity are examined from ERA-5 reanalysis data and RWP measurements. The contribution of each feature variable to the performance of the RF model in independent tests is also discussed here. The results show that the usage of RWP observational data as the RF model input tends to result in better performance in rainfall/non-rainfall forecast 30 min in advance of rainfall onset than using the ERA-5 data as inputs. For the rainfall grade forecast, the divergence and vorticity that were estimated from the RWP measurements at 800 hPa show importance in improving the model performance in heavy and moderate rain forecasts. This indicates that the atmospheric dynamic variable measurements from RWP have great potential to improve the prediction skill of convection with the aid of a machine learning model. [ABSTRACT FROM AUTHOR]

Landslide susceptibility assessment is an important means of helping to reduce and manage landslide risk. The existing studies, however, fail to examine the spatially varying relationships between landslide susceptibility and its explanatory factors. This paper investigates the spatial variation in such relationships in Liangshan, China, leveraging a spatially explicit model, namely, geographical random forest (GRF). By comparing with random forest (RF), we found that GRF achieves a higher performance with an AUC of 0.86 due to its consideration of the spatial heterogeneity among variables. GRF also provides a higher-quality landslide susceptibility map than RF by correctly placing 92.35% of the landslide points in high-susceptibility areas. The local feature importance derived from GRF allows us to understand that the impact of conditioning factors varies across space, which can provide implications for policy development by local governments to place different levels of attention on different conditioning factors in specific counties to prevent and mitigate landslides. To account for the spatial dependence among the data in the model performance assessment, we use spatial cross-validation (CV) to split the data into subsets spatially rather than randomly for model training and testing. The results show that spatial CV can effectively address the over-optimistic bias in model error evaluation. [ABSTRACT FROM AUTHOR]

Disclosure is the soul of supervision, the bridge between companies and investors, and can help public fully understand companies' business operations. The disclosure quality of listed companies' annual reports affects security market efficiency and investor rights protection. Currently, there are issues including but not limited to contradictory information and avoidance of important problems. Moreover, the qualities of disclosure reports from different companies vary greatly. This paper is to focus on annual reports of listed companies, dividing them into two categories based on the assessment results of disclosure by the Shenzhen Stock Exchange, and conducting the research on textual characteristics. Firstly, three characteristic indicators, tone, readability, and file size of each annual report are analyzed and compared, with validation. To explore the impact of characteristic indicators on text classification, these three indicators are introduced into text models, constructing comprehensive models. To achieve dimension reduction for training models, feature selection is performed by using Chi-square statistics. Different lengths keywords dictionaries are constructed. In conclusion, prediction performance of classifier models can be improved or maintained after introducing indicators, with the random forest model having the best improvement after introduction of indicators. [ABSTRACT FROM AUTHOR]

Multi-source data is widely used in the field of energy future prices forecasting, the improvement of forecast ability and data screening are becoming the focus of current research. In this paper, two tree-based models (namely, Random Forest and XGBoost model) are employed to predict China's crude oil future prices. The empirical analysis confirms that Random Forest and XGBoost model have superior prediction performances than benchmark and the XGBoost model performs best. An important finding is that there is a time gap between investor information search and processing because the prediction performance within the time lags is obviously superior than that of the current period. [ABSTRACT FROM AUTHOR]