Learning disability and leadership: Becoming an effective leader

This study investigates learning disability (LD) as an individual‐differences variable predicting leadership emergence, role occupancy, and effectiveness. We hypothesize that individuals with LD are less likely to occupy leadership roles, and that informal group processes (leadership emergence) will mediate the relationship between LD and leadership role occupancy. We also hypothesized that, among leaders promoted and selected for leadership training, there would be a negative relationship between LD and effective leadership. We first checked for LD in a sample of 1076 soldiers, measuring cognitive ability with a geometric‐analogies test as a control. Some months later, during the soldiers' basic training, we measured leadership emergence. We then identified those who were selected for leadership training, recording, and measuring their effectiveness according to supervisory and peer evaluations. Leadership emergence was found to mediate the negative relationship between LD and leadership role occupancy. There were no significant differences among leaders (n = 308) with and without LD in regard to leadership effectiveness. Copyright © 2013 John Wiley & Sons, Ltd.     

Risk aversion and adaptive management: Insights from a multi-armed bandit model of invasive species risk

This article explores adaptive management (AM) for decision-making under environmental uncertainty. In the context of targeting invasive species inspections of agricultural imports, I find that risk aversion increases the relative value of AM and can increase the rate of exploratory action. While calls for AM in natural resource management are common, many analyses have identified modest gains from this approach. I analytically and numerically examine the distribution of outcomes from AM under risk neutrality and risk aversion. The inspection decision is framed as a multi-armed bandit problem and solved using the Lagrangian decomposition method. Results show that even when expected gains are modest, asymmetry in the distribution of outcomes has important implications. Notably, AM can serve to buffer against large losses, even if the most likely outcome is a small loss.     

Social learning for sustainability: supporting sustainable business in Brazil regarding multiple social actors,relationships and interests

Finding solutions for complex environmental, economic and social issues in organisations relies on coordinated actions among several social actors that are involved in the sustainable development web and demands that they learn new business methods. In this scenario, which involves multiple social actors, relationships, contexts and interests, social learning has emerged as a hybrid approach to resolving complex environmental and social problems. Such an approach marks a different situation for organisations whose focus is not only on supporting such problem resolution but also on transforming such crises into opportunities to generate sustainable products and services. Thus, this study discusses how one of the largest companies in Latin America’s chemical segment formed a network with 23 social actors and is socially learning to deal with the dialectic between return on investments for its shareholders and benefits for its stakeholders. Research was conducted based on Boje’s narrative analysis. The data was built through interviews, informal conversations, textual and audio-visual documents and non-participant observation. This paper discusses the concept and describes the social learning process for sustainability (in this case, for sustainable agriculture) from the viewpoint of a for-profit organisation. In this manner, this study contributes to strengthening the connection between social learning and sustainability.     

Automated conservation assessment of the orchid family with deep learning

International Union for Conservation of Nature (IUCN) Red List assessments are essential for prioritizing conservation needs but are resource intensive and therefore available only for a fraction of global species richness. Automated conservation assessments based on digitally available geographic occurrence records can be a rapid alternative, but it is unclear how reliable these assessments are. We conducted automated conservation assessments for 13,910 species (47.3% of the known species in the family) of the diverse and globally distributed orchid family (Orchidaceae), for which most species (13,049) were previously unassessed by IUCN. We used a novel method based on a deep neural network (IUC-NN). We identified 4,342 orchid species (31.2% of the evaluated species) as possibly threatened with extinction (equivalent to IUCN categories critically endangered [CR], endangered [EN], or vulnerable [VU]) and Madagascar, East Africa, Southeast Asia, and several oceanic islands as priority areas for orchid conservation. Orchidaceae provided a model with which to test the sensitivity of automated assessment methods to problems with data availability, data quality, and geographic sampling bias. The IUC-NN identified possibly threatened species with an accuracy of 84.3%, with significantly lower geographic evaluation bias relative to the IUCN Red List and was robust even when data availability was low and there were geographic errors in the input data. Overall, our results demonstrate that automated assessments have an important role to play in identifying species at the greatest risk of extinction.     

基于RF-ELM模型的边坡稳定性预测研究

为实现对边坡稳定性的有效预测,将极限学习机算法与旋转森林算法相结合,并依据影响边坡稳定性的六项重要因素,建立了边坡稳定性预测的RF-ELM预测模型。该模型是以极限学习机算法为基分类器,以旋转森林算法为框架的集成学习模型,利用UCI数据库中三组数据集验证了该集成模型确实提高了ELM的预测性能。将RF-ELM模型应用于边坡稳定性的预测问题中,结合39组工程实例数据进行预测实验,结果表明该模型具有较高的预测精度,可有效的对边坡稳定性进行预测。     

Evidence-based guidelines for automated conservation assessments of plant species

Assessing species’ extinction risk is vital to setting conservation priorities. However, assessment endeavors, such as those used to produce the IUCN Red List of Threatened Species, have significant gaps in taxonomic coverage. Automated assessment (AA) methods are gaining popularity to fill these gaps. Choices made in developing, using, and reporting results of AA methods could hinder their successful adoption or lead to poor allocation of conservation resources. We explored how choice of data cleaning type and level, taxonomic group, training sample, and automation method affect performance of threat status predictions for plant species. We used occurrences from the Global Biodiversity Information Facility (GBIF) to generate assessments for species in 3 taxonomic groups based on 6 different occurrence-based AA methods. We measured each method's performance and coverage following increasingly stringent occurrence cleaning. Automatically cleaned data from GBIF performed comparably to occurrence records cleaned manually by experts. However, all types of data cleaning limited the coverage of AAs. Overall, machine-learning-based methods performed well across taxa, even with minimal data cleaning. Results suggest a machine-learning-based method applied to minimally cleaned data offers the best compromise between performance and species coverage. However, optimal data cleaning, training sample, and automation methods depend on the study group, intended applications, and expertise.     

湖南空气质量预报中的数据预处理和特征工程

为提高空气质量预报的准确率,建立了融合气象和环境观测资料、结合机器学习和数值天气预报,且预测时效较长、预测精度较高的机器学习模型库。以湖南6个城市(长沙、株洲、湘潭、益阳、常德、岳阳)的空气质量预报为例,将数据预处理、特征工程方法运用到模型之中,得出以下几点结论:①数据预处理工作包括样本收集、数据清洗、缺失值处理、异常值剔除等,对提高模型预测稳定性帮助很大。②点、线、面的特征组合有助于完整地描述污染物的生消过程。引入传输指数后,株洲市模型对传输型污染过程的预测性能得到明显提高,对轻度、中度、重度污染的分类准确度分别提升了23.6%、16.6%、30.0%。引入静稳指数后,长沙市模型PM_2.5浓度测试的相关系数由0.938提升至0.959,均方根误差由10.33下降至8.46,且模型对中度以上污染天气的极值预报结果更接近实况;益阳市模型在高浓度样本预测中存在的系统性偏低现象得到改善,对轻度以上污染天气的预报结果得到较大矫正。③随机森林的特征重要性排序功能可以大幅度减少特征的数量,使得模型的可解释性和稳定性增强。     

深度学习在裸地扬尘源监测中的应用研究

裸地是扬尘的重要来源,施工建设过程中形成的裸地极易在大风天气作用下造成扬尘污染。因此,快速、有效地定位裸地位置,并确认其管控措施落实情况,对于开展裸地扬尘源监管具有重要意义。基于高分辨率遥感监测数据,结合人工解译裸地扬尘源数据集,以北京市大兴区为例,利用深度学习方法对裸地和防尘网覆盖裸地进行分类识别。同时,利用颜色匹配法对大兴区防尘网覆盖裸地进行识别,横向评估深度学习方法的识别精度。结果显示：深度学习方法对防尘网覆盖裸地的识别精度达97%,对裸地的识别精度达61%;颜色匹配法对防尘网覆盖裸地的识别精度达85%。防尘网覆盖裸地的颜色特征鲜明,深度学习方法和颜色匹配法对防尘网覆盖裸地的识别精度都在85%以上。深度学习方法对于面积大于2 000 m²的图斑有着较好的识别精度。深度学习方法可以提高裸地遥感解译的效率,实现规范化图像识别,可以作为人工判读的辅助手段。在实际应用中,可通过进一步积累样本来增强模型性能。深度学习方法适用于裸地扬尘源线索快速发现、工地防尘网措施落实情况快速检测等场景。     

大气环境分类的SVM联合决策方法

目的 实现城市大气环境的精准快速归类预测。方法 基于支持向量机（SVM）构建多分类问题的联合决策算法,将大量城市环境因素数据的主成分聚类结果作为输入,通过机器学习训练,组建大气环境的SVM联合决策模型。结果 该模型根据大气环境因素将数据集91个城市划分为9类,其中河内与海防环境相似度最高,巴东与格尔木差异最大。9个SVM二分类器组建的联合决策模型通过逐点预测在主成分数据空间形成了分区预测云图。结论 SVM联合决策模型可实现城市环境的快速分类辨识,分类预测结果精度高于95%。     

基于机器学习的恐怖袭击事件后果预测方法研究

恐怖袭击事件通常会造成严重的人员伤亡、财产损失和社会影响,针对在不同场景下发生恐怖袭击所造成的后果进行预测是目前应对恐怖袭击事件急需解决的问题之一。利用多源数据,首先基于随机森林算法对恐怖袭击事件是否造成死伤进行分类预测,进而基于岭回归算法预测事件造成的具体死伤人数。研究结果表明:随机森林在测试集上对有死伤事件的召回率达到0.85,岭回归预测死亡和受伤人数的平均绝对误差分别小于1人和2人。研究结果可为反恐资源配置优化、预防恐怖袭击事件和减少其造成的损害提供辅助决策支持。