Alhagi sparsifolia Shap. （Fabaceae） is a spiny, perennial herb. The species grows in the salinized, arid regions in North China. This study investigated the response characteristics of the root growth and the dis- t...Alhagi sparsifolia Shap. （Fabaceae） is a spiny, perennial herb. The species grows in the salinized, arid regions in North China. This study investigated the response characteristics of the root growth and the dis- tribution of one-year-old A. sparsifolia seedlings to different groundwater depths in controlled plots. The eco- logical adaptability of the root systems of A. sparsifolia seedlings was examined using the artificial digging method. Results showed that： （1） A. sparsifolia seedlings adapted to an increase in groundwater depth mainly through increasing the penetration depth and growth rate of vertical roots. The vertical roots grew rapidly when soil moisture content reached 3%-9%, but slowly when soil moisture content was 13%-20%. The vertical roots stopped growing when soil moisture content reached 30% （the critical soil moisture point）. （2） The morphological plasticity of roots is an important strategy used by A. sparsifolia seedlings to obtain water and adapt to dry soil conditions. When the groundwater table was shallow, horizontal roots quickly expanded and tillering increased in order to compete for light resources, whereas when the groundwater table was deeper, vertical roots developed quickly to exploit space in the deeper soil layers. （3） The decrease in groundwater depth was probably respon- sible for the root distribution in the shallow soil layers. Root biomass and surface area both decreased with soil depth. One strategy of A. sparsifolia seedlings in dealing with the increase in groundwater depth is to increase root biomass in the deep soil layers. The relationship between the root growth/distribution of A. sparsifolia and the depth of groundwater table can be used as guidance for harvesting A. sparsifolia biomass and managing water resources for forage grasses. It is also of ecological significance as it reveals how desert plants adapt to arid environments.展开更多
【目的】对可解释机器学习方法及其在信息资源管理领域的应用研究进行梳理和总结,发现不足并做出展望。【文献范围】使用可解释机器学习的相关关键词构建检索式,在中国知网和Web of Science等平台中检索,根据纳入排除标准,共获取44篇相...【目的】对可解释机器学习方法及其在信息资源管理领域的应用研究进行梳理和总结,发现不足并做出展望。【文献范围】使用可解释机器学习的相关关键词构建检索式,在中国知网和Web of Science等平台中检索,根据纳入排除标准,共获取44篇相关文献进行评述。【方法】从机器学习流程出发,构建一般性的可解释机器学习框架,重点梳理可解释机器学习方法分类,然后对可解释机器学习在信息资源管理领域的应用现状进行归纳总结。【结果】一般性的可解释机器学习框架包含事前解释、可解释模型以及事后解释三个不同的模块;事后可解释方法在健康信息学、网络舆情、科学计量学以及社交网络用户行为等领域具有广泛的应用,其中常用的方法为SHAP和特征重要性分析;现有研究存在应用方法单一和融合不足、因果关系探究不够、针对多源异构数据的解释不足以及领域应用有待拓宽等问题。【局限】本文重点关注可解释机器学习的应用及存在的不足,未对算法原理进行深入阐述。【结论】未来研究应加强可解释机器学习方法的融合使用,探究基于因果机器学习的可解释机器学习,引入面向多源异构数据的可解释机器学习方法,拓宽在信息推荐、信息检索和信息计量等多个领域的应用。展开更多
城市轨道交通作为低能耗、少污染、具有可持续属性的公共交通类型之一,其对沿线城市发展、居民生产生活产生深远影响。中国城市轨道交通建设目前仍处于高速发展阶段,部分站点周边地区面临空间利用率不匹配、潮汐客流趋势加重等问题。城...城市轨道交通作为低能耗、少污染、具有可持续属性的公共交通类型之一,其对沿线城市发展、居民生产生活产生深远影响。中国城市轨道交通建设目前仍处于高速发展阶段,部分站点周边地区面临空间利用率不匹配、潮汐客流趋势加重等问题。城市轨道交通站点周边地区的城市空间规划需关注城市居民的活动特征,以提升站点地区城市空间全时段活力。以南京市中心城区内轨道交通站点周边地区为例,基于城市空间开放数据、实地踏勘调研、互联网移动定位服务(location based service,LBS)数据,采集统计与评价建成环境现状与居民活动特征数据,并运用梯度提升决策树与SHAP(Shapley addictive explanation)解释分析站点地区建成环境与居民活动的非线性关系及建成环境要素之间的交互作用,在此基础上提出建成环境要素适宜区间及协同优化条件,为城市轨道交通站点周边地区空间规划与优化提供建议。展开更多
基金supported by the Knowledge Innovation Program of the Chinese Academy of Sciences (KZCX2-EW-316)the National Natural Science Foundation of China (31070477,30870471)the West Light Foundation of the Chinese Academy of Sciences (XBBS201105)
文摘Alhagi sparsifolia Shap. （Fabaceae） is a spiny, perennial herb. The species grows in the salinized, arid regions in North China. This study investigated the response characteristics of the root growth and the dis- tribution of one-year-old A. sparsifolia seedlings to different groundwater depths in controlled plots. The eco- logical adaptability of the root systems of A. sparsifolia seedlings was examined using the artificial digging method. Results showed that： （1） A. sparsifolia seedlings adapted to an increase in groundwater depth mainly through increasing the penetration depth and growth rate of vertical roots. The vertical roots grew rapidly when soil moisture content reached 3%-9%, but slowly when soil moisture content was 13%-20%. The vertical roots stopped growing when soil moisture content reached 30% （the critical soil moisture point）. （2） The morphological plasticity of roots is an important strategy used by A. sparsifolia seedlings to obtain water and adapt to dry soil conditions. When the groundwater table was shallow, horizontal roots quickly expanded and tillering increased in order to compete for light resources, whereas when the groundwater table was deeper, vertical roots developed quickly to exploit space in the deeper soil layers. （3） The decrease in groundwater depth was probably respon- sible for the root distribution in the shallow soil layers. Root biomass and surface area both decreased with soil depth. One strategy of A. sparsifolia seedlings in dealing with the increase in groundwater depth is to increase root biomass in the deep soil layers. The relationship between the root growth/distribution of A. sparsifolia and the depth of groundwater table can be used as guidance for harvesting A. sparsifolia biomass and managing water resources for forage grasses. It is also of ecological significance as it reveals how desert plants adapt to arid environments.
文摘【目的】对可解释机器学习方法及其在信息资源管理领域的应用研究进行梳理和总结,发现不足并做出展望。【文献范围】使用可解释机器学习的相关关键词构建检索式,在中国知网和Web of Science等平台中检索,根据纳入排除标准,共获取44篇相关文献进行评述。【方法】从机器学习流程出发,构建一般性的可解释机器学习框架,重点梳理可解释机器学习方法分类,然后对可解释机器学习在信息资源管理领域的应用现状进行归纳总结。【结果】一般性的可解释机器学习框架包含事前解释、可解释模型以及事后解释三个不同的模块;事后可解释方法在健康信息学、网络舆情、科学计量学以及社交网络用户行为等领域具有广泛的应用,其中常用的方法为SHAP和特征重要性分析;现有研究存在应用方法单一和融合不足、因果关系探究不够、针对多源异构数据的解释不足以及领域应用有待拓宽等问题。【局限】本文重点关注可解释机器学习的应用及存在的不足,未对算法原理进行深入阐述。【结论】未来研究应加强可解释机器学习方法的融合使用,探究基于因果机器学习的可解释机器学习,引入面向多源异构数据的可解释机器学习方法,拓宽在信息推荐、信息检索和信息计量等多个领域的应用。
文摘城市轨道交通作为低能耗、少污染、具有可持续属性的公共交通类型之一,其对沿线城市发展、居民生产生活产生深远影响。中国城市轨道交通建设目前仍处于高速发展阶段,部分站点周边地区面临空间利用率不匹配、潮汐客流趋势加重等问题。城市轨道交通站点周边地区的城市空间规划需关注城市居民的活动特征,以提升站点地区城市空间全时段活力。以南京市中心城区内轨道交通站点周边地区为例,基于城市空间开放数据、实地踏勘调研、互联网移动定位服务(location based service,LBS)数据,采集统计与评价建成环境现状与居民活动特征数据,并运用梯度提升决策树与SHAP(Shapley addictive explanation)解释分析站点地区建成环境与居民活动的非线性关系及建成环境要素之间的交互作用,在此基础上提出建成环境要素适宜区间及协同优化条件,为城市轨道交通站点周边地区空间规划与优化提供建议。
