Prediction of pullout interaction coefficient of geogrids by extreme gradient boosting model

Geogrids embedded in fill materials are checked against pullout failure through standard pullout testing methodology. The test determines the pullout interaction coefficient which is critical in fixing the embedment length of geogrids in mechanically stabilized earth walls. This paper proposes prediction of pullout interaction coefficient using data driven machine learning regression algorithms. The study primarily focusses on using extreme gradient boosting (XGBoost) method for prediction. A data set containing 220 test results from the literature has been used for training and testing. Predicted results of XGBoost have been compared with the results of random forest (RF) ensemble learning based algorithm. The predictions of XGBoost model indicates 85% accuracy and that of RF model shows 77% accuracy, indicating significantly superior and robust prediction through XGBoost above RF model. The importance analysis indicates that normal stress is the most significant factor that influences the pullout interaction coefficients. Subsequently pullout tests have been performed on geogrid embedded in four different fill materials at three normal stresses. The proposed XGBoost model gives 90% accuracy in prediction of pullout interaction coefficient compared to laboratory test results. Finally, an open-source graphical user interface based on the XGBoost model has been created for preliminary estimation of the pullout interaction coefficient of geogrid at different test conditions.     

Analyzing growth and mortality in a subtropical urban forest ecosystem

Information on urban tree growth, mortality and in-growth is currently being used to estimate urban forest structure changes and ecosystem services such as carbon sequestration. This study reports on tree diameter growth and mortality in 65 plots distributed among four land use categories, which were established in 2005/2006 in Gainesville, Florida, USA and were re-measured in 2009. Models for mortality and in-growth models were developed by grouping species into hardwoods and softwoods. Annual change in tree diameter at breast height growth was analyzed using three tree species groups based on potential height and longevity. Additionally, the four most common tree species in the study area were modeled to explore factors affecting tree growth. The average annual mortality rate in the city was 9.97%. Trees located in Institutional land use/land cover (LULC) had the highest annual mortality rate (19.2%/yr), and commercial had the lowest (3.1%/yr). Overall, growth rates for the study area (0.70 cm/yr) and residential LULC (0.80 cm/yr) were comparable to other studies. Growth rates for trees in forested areas were higher (0.56 cm/yr) than those previously reported. Individual species-level growth rates such as those for Juniperus virginiana (1.24 cm/yr) and Quercus virginiana (1.08 cm/yr) were different than other species values reported in other studies. Maintenance activities, site conditions, soil properties, tree characteristics, and LULC significantly influenced urban tree growth, mortality, and in-growth. Results can be used to better understand urban forest ecosystem structure and services in medium sized, subtropical cities and to make better decisions regarding planting and maintenance strategies.     

基于 Windows Mobile 的森林资源连续清查软件研发

随着数字林业的发展，无纸化外业数据采集成为一种发展趋势。移动数据采集软件能够根据样地中心点的坐标进行导航，为调查人员寻找样地带来方便；能够快速、准确采集森林资源的各种属性信息及空间位置信息数据；数据检核模块能够自动检查作业员在数据采集中出现的逻辑关系错误，并给出提示，便于调查人员直接修改，保证数据成果的质量。本文以第八次全国森林资源连续清查为例，从野外数据采集到内业数据汇总，介绍该软件的设计思路和主要功能，充分发挥其在资源调查中的优势，更好地为现代林业服务。     

An analytical comparison of four approaches to modelling the daily variability of solar irradiance using meteorological records

Temporal solar variability significantly affects the integration of solar power systems into the grid. It is thus essential to predict temporal solar variability, particularly given the increasing popularity of solar power generation globally. In this paper, the daily variability of solar irradiance at four sites across Australia is quantified using observed time series of global horizontal irradiance for 2003–2012. It is shown that the daily variability strongly depends on sky clearness with generally low values under a clear or overcast condition and high values under an intermittent cloudiness condition. Various statistical techniques are adopted to model the daily variability using meteorological variables selected from the ERA-Interim reanalysis as predictors. The nonlinear regression technique (i.e. random forest) is demonstrated to perform the best while the performance of the simple analog method is only slightly worse. Among the four sites, Alice Springs has the lowest daily variability index on average and Rockhampton has the highest daily variability index on average. The modelling results of the four sites produced by random forest have a correlation coefficient of above 0.7 and a median relative error around 40%. While the approach of statistical downscaling from a large spatial domain has been applied for other problems, it is shown in this study that it generally suffices to use only the predictors at a single near point for the problem of solar variability. The relative importance of the involved meteorological variables and the effects of clearness on the modelling of the daily variability are also explored.     

Assessment of potential bioenergy from coppice forests trough the integration of remote sensing and field surveys

A spatially explicit knowledge of forest resources is essential to support the sustainable use of wood as a fuel for producing energy (firewood).This paper describes the integrated use of remotely sensed data and sample based forest inventories to derive a biomass map for coppice forest, resulted estimated potential biomass available is contrasted with local domestic consumptions at the municipality level. The test was carried out in an environmentally and socially homogeneous district of Apennine Mountains (Alto Molise, south-central Italy) coupling multispectral high resolution Landsat 7 ETM+ satellite imagery and a local forest inventory trough the application of the non-parametric estimation procedure k-Nearest Neighbours (k-NN). Several forest management scenarios were applied in order to evaluate their impact on the potential availability of firewood from coppice forests.The paper introduces data and methods used and presents the achieved results both in terms of the accuracy of the biomass map produced by k-NN and of the relationship between the potential availability and demand for firewood.These results demonstrated that k-NN is able to estimate the biomass of coppice forest in the test area with an accuracy level comparable with recent similar application of k-NN carried out in Boreal regions (RMSE of 25.6%).The application of different forest management scenarios have a significant impact on local estimated firewood balance between potential supply from coppice forests and demand for domestic consumption, depending of the scenarios the net balance changed up to 84%.     

Forest Fire Detection Using Artificial Neural Network Algorithm Implemented in Wireless Sensor Networks

A forest fire is a severe threat to forest resources and human life. In this paper, we propose a forest?fire detection system that has an artificial neural network algorithm implemented in a wireless s...     

基于大数据挖掘分析的垃圾短信治理方案

本文简要介绍了垃圾短信对社会带来的负面影响,分析了目前垃圾短信治理体系存在的不足,提出了基于大数据挖掘分析的垃圾短信治理方案:首先构建垃圾短信用户识别指标体系,通过引入随机森林模型,对垃圾短信用户进行识别和有效治理,最后运用平台对垃圾短信用户进行精细化管理。     

Trajectory Analysis of Forest Cover Change in the Tropical Dry Forest of Burkina Faso,West Africa

Forest cover decline is one of the most important environmental issues in the tropics. The present study was carried out in Burkina Faso, West Africa, and aimed at assessing the trajectories of forest cover change and measuring landscape metrics of the trajectory classes in order to better understand the processes of change. Landsat and ASTER images acquired over a period of 30 years were used for cover change detection and the Fragstats package was used to compute landscape metrics with five unifying change classes. Results showed a substantial increase in cropland with concurrent decline in forest cover. Deforestation represented 63% of the Percentage of Landscape (PLAND) in 2006, while reforestation accounted for only 28%. Both of these classes had high Normalized Landscape Shape Index (NLSI) values, indicating that they were present as scattered small patches. The old cultivation (30-year permanent cropland) was aggregated (IJI ≈ 0) while deforestation exhibited highly interspersed patches. The old forest and old cultivation presented lower Area Weighted Fractal Dimension Index (FRAC_AM), but deforestation and reforestation had the higher FRAC_AM. These results confirmed that there was a high level of deforestation and fragmentation in southern Burkina Faso and justify the need for a proper management plan to ensure the sustainable use of forest resources.     

A Novel Feature Extraction Method with Feature Selection to Identify Golgi-Resident Protein Types from Imbalanced Data

The Golgi Apparatus (GA) is a major collection and dispatch station for numerous proteins destined for secretion, plasma membranes and lysosomes. The dysfunction of GA proteins can result in neurodegenerative diseases. Therefore, accurate identification of protein subGolgi localizations may assist in drug development and understanding the mechanisms of the GA involved in various cellular processes. In this paper, a new computational method is proposed for identifying cis-Golgi proteins from trans-Golgi proteins. Based on the concept of Common Spatial Patterns (CSP), a novel feature extraction technique is developed to extract evolutionary information from protein sequences. To deal with the imbalanced benchmark dataset, the Synthetic Minority Over-sampling Technique (SMOTE) is adopted. A feature selection method called Random Forest-Recursive Feature Elimination (RF-RFE) is employed to search the optimal features from the CSP based features and g-gap dipeptide composition. Based on the optimal features, a Random Forest (RF) module is used to distinguish cis-Golgi proteins from trans-Golgi proteins. Through the jackknife cross-validation, the proposed method achieves a promising performance with a sensitivity of 0.889, a specificity of 0.880, an accuracy of 0.885, and a Matthew’s Correlation Coefficient (MCC) of 0.765, which remarkably outperforms previous methods. Moreover, when tested on a common independent dataset, our method also achieves a significantly improved performance. These results highlight the promising performance of the proposed method to identify Golgi-resident protein types. Furthermore, the CSP based feature extraction method may provide guidelines for protein function predictions.     

基于深度森林的热轧带钢凸度预测模型

在热轧现场轧制规格切换或工况异常的情况下板凸度控制模型偏差较大,为了提高模型精度,提出了一种基于深度森林的热轧带钢凸度预测模型。深度森林模型融合了集成学习和深度学习的思想,采用了多粒度扫描增加数据特征多样性,采用级联森林逐层处理,使得模型具备强大数据拟合能力。将热轧数据经前期预处理导入模型,并对模型参数进行了网格搜索寻优,对比随机森林模型,深度森林模型的效果更优。基于深度森林的热轧带钢凸度预测模型得到了MSE值为6.537,MAE值为1.587,MAPE值为2.903%和R值为0.985的预测性能。