可持续发展目标(Sustainable Development Goals,SDGs)是联合国用于指导全球可持续性发展工作的新议程。各项目标之间的关联性,尤其是3级指标数据的相关性分析和研究,对于有效推动可持续发展目标的具体实施具有重要意义。本文基于《可...可持续发展目标(Sustainable Development Goals,SDGs)是联合国用于指导全球可持续性发展工作的新议程。各项目标之间的关联性,尤其是3级指标数据的相关性分析和研究,对于有效推动可持续发展目标的具体实施具有重要意义。本文基于《可持续发展报告2020》中联合国全体193个成员国的2000–2017年的3级指标时间序列数据,经过数据的预处理,并依据算法的相关性覆盖程度的大小,从Pearson、Spearman、Kendall和MIC 4个算法中遴选出MIC算法和Spearman算法,计算得到193个成员国2000–2017年总计62项可利用指标项、1766个指标对的相关系数和相关方向。本数据集可以作为联合国评估可持续发展目标未来实现的分析数据,也可以作为各国监测指标完成情况、制定相关政策的重要参考依据。展开更多
Well-managed forests are major sources of livelihoods for the fringed communities. However, the remoteness, inaccessibility of most forested areas coupled with conflicts from adjacent forest communities, who often dep...Well-managed forests are major sources of livelihoods for the fringed communities. However, the remoteness, inaccessibility of most forested areas coupled with conflicts from adjacent forest communities, who often depend on it for livelihood is a daunting task in implementing conservation, viz-a-viz the Sustainable Development Goals. The Mt Oku forest is a unique, remote but represents novelty in forest management in remote areas in Cameroon, with devolution of management rights. The forest is well noted for its high level of endemism. This study is focused on the legal, institutional, socio-economic and regulatory framework put in place, for appropriate conservation and livelihood sustenance as forest management rights were devolved to the local community. A multidimensional framework guiding the development of testable hypothesis that assesses the relationship between the forest users’ activities and forest degradation, which have a multiplier effect on the SDGs, was used. The alternate livelihood options/strategies and benefits after the institution of the Forest Management Project (FMP) was examined in randomly selected frontline and secondary villages. Selected Participatory Rural Appraisal (PRA) tools and registers from FMIs aided relevant data collection. The results indicated a significant relationship (χ2 = 0.65.4, p = 0.00) between the activities of forest users and forest degradation. Adopted alternate livelihood strategies/options include, direct employment, tourist guides, bee farming, agricultural intensification, agroforestry, capacity building for skill acquisition, selective exploitation and sales of Prunus spp. Forest regeneration strategies ranged from, forest guards, removal of exotic species, forest enrichment, raising nurseries, fire tracing, etc. It is concluded that giving greater access and ownership of forest to the local community in the Mount Oku Region, led to transparency, accountability and social stability, which contributed tremendously both to the recovery and conservation of 展开更多
【目的】联合国可持续发展目标(Sustainable Development Goals,SDGs)已经成为全球最重要的可持续发展问题。然而,SDGs指标相关数据高缺失率的现状严重影响了联合国对各国可持续发展目标实行过程的有效监测。研究如何对SDGs中的相关缺...【目的】联合国可持续发展目标(Sustainable Development Goals,SDGs)已经成为全球最重要的可持续发展问题。然而,SDGs指标相关数据高缺失率的现状严重影响了联合国对各国可持续发展目标实行过程的有效监测。研究如何对SDGs中的相关缺失数据进行补全具有重大的技术挑战,也对鞭策各国完成可持续发展目标具备重大意义。【方法】本文提出了一种融合MIC(最大信息系数)进行特征选择的迁移学习方法TLM(—种融合最大信息系数和迁移学习的方法),其能通过其它公开数据为目标变量构造特征,并联合相关回归技术建立数据预测模型,以达到对目标变量的缺失值进行预测的目的。【结果】本文以特定国家中SDGs指标3.2.1的数据集为例,使用TLM方法对目标变量的缺失值进行预测并补全,验证了TLM方法的有效性。【局限】由于影响SDGs指标的波动因素众多,因此,探索更多相关性分析方法并结合TLM方法对缺失值进行更加精确的预测是今后进一步研究的重点方向。【结论】结合了MIC和迁移学习的TLM方法能提升数据预测的准确率,可为SDGs相关领域工作者在处理数据缺失问题时提供重要的参考价值。展开更多
目的分析2010-2017年我国31个省市道路交通事故发生数和死伤情况,预测到2020年各省市实现可持续发展目标(sustainable development goals,SDGs)中设定的死伤人数较2015年减半的情况。方法分析31个省市道路交通事故现状,利用趋势外推法预...目的分析2010-2017年我国31个省市道路交通事故发生数和死伤情况,预测到2020年各省市实现可持续发展目标(sustainable development goals,SDGs)中设定的死伤人数较2015年减半的情况。方法分析31个省市道路交通事故现状,利用趋势外推法预测2020年能否实现SDGs目标。结果2010-2017年广东、江苏、浙江、山东的交通事故发生数、死伤数呈下降趋势但仍居前列。2017年死亡率前两位是北京和贵州,受伤率贵州和天津居高。预测显示,到2020年全国道路交通事故的死亡和受伤人数分别增加了17%和1.3%,难以实现减半的目标;预计2020年湖南的死伤人数分别降至2015年的50.7%和65.3%,能够达到目标;上海的受伤人数降低83.3%,但死亡人数只降低34.5%,与目标有差距;其余省市难以达到目标。预计到2020年11个省市的死亡人数将增加,其中北京、吉林、江西、湖北和贵州增长最为明显;有8个省市的受伤人数呈增长趋势,其中吉林、江西、湖北和贵州四省增幅明显。结论除湖南省外,全国和各省市均难以达到SDG s目标,可以根据预测结果和实际现状制定有针对性的科学防控策略。湖南和上海在交通事故防控方面的措施有效,值得借鉴。展开更多
文摘可持续发展目标(Sustainable Development Goals,SDGs)是联合国用于指导全球可持续性发展工作的新议程。各项目标之间的关联性,尤其是3级指标数据的相关性分析和研究,对于有效推动可持续发展目标的具体实施具有重要意义。本文基于《可持续发展报告2020》中联合国全体193个成员国的2000–2017年的3级指标时间序列数据,经过数据的预处理,并依据算法的相关性覆盖程度的大小,从Pearson、Spearman、Kendall和MIC 4个算法中遴选出MIC算法和Spearman算法,计算得到193个成员国2000–2017年总计62项可利用指标项、1766个指标对的相关系数和相关方向。本数据集可以作为联合国评估可持续发展目标未来实现的分析数据,也可以作为各国监测指标完成情况、制定相关政策的重要参考依据。
文摘Well-managed forests are major sources of livelihoods for the fringed communities. However, the remoteness, inaccessibility of most forested areas coupled with conflicts from adjacent forest communities, who often depend on it for livelihood is a daunting task in implementing conservation, viz-a-viz the Sustainable Development Goals. The Mt Oku forest is a unique, remote but represents novelty in forest management in remote areas in Cameroon, with devolution of management rights. The forest is well noted for its high level of endemism. This study is focused on the legal, institutional, socio-economic and regulatory framework put in place, for appropriate conservation and livelihood sustenance as forest management rights were devolved to the local community. A multidimensional framework guiding the development of testable hypothesis that assesses the relationship between the forest users’ activities and forest degradation, which have a multiplier effect on the SDGs, was used. The alternate livelihood options/strategies and benefits after the institution of the Forest Management Project (FMP) was examined in randomly selected frontline and secondary villages. Selected Participatory Rural Appraisal (PRA) tools and registers from FMIs aided relevant data collection. The results indicated a significant relationship (χ2 = 0.65.4, p = 0.00) between the activities of forest users and forest degradation. Adopted alternate livelihood strategies/options include, direct employment, tourist guides, bee farming, agricultural intensification, agroforestry, capacity building for skill acquisition, selective exploitation and sales of Prunus spp. Forest regeneration strategies ranged from, forest guards, removal of exotic species, forest enrichment, raising nurseries, fire tracing, etc. It is concluded that giving greater access and ownership of forest to the local community in the Mount Oku Region, led to transparency, accountability and social stability, which contributed tremendously both to the recovery and conservation of
文摘【目的】联合国可持续发展目标(Sustainable Development Goals,SDGs)已经成为全球最重要的可持续发展问题。然而,SDGs指标相关数据高缺失率的现状严重影响了联合国对各国可持续发展目标实行过程的有效监测。研究如何对SDGs中的相关缺失数据进行补全具有重大的技术挑战,也对鞭策各国完成可持续发展目标具备重大意义。【方法】本文提出了一种融合MIC(最大信息系数)进行特征选择的迁移学习方法TLM(—种融合最大信息系数和迁移学习的方法),其能通过其它公开数据为目标变量构造特征,并联合相关回归技术建立数据预测模型,以达到对目标变量的缺失值进行预测的目的。【结果】本文以特定国家中SDGs指标3.2.1的数据集为例,使用TLM方法对目标变量的缺失值进行预测并补全,验证了TLM方法的有效性。【局限】由于影响SDGs指标的波动因素众多,因此,探索更多相关性分析方法并结合TLM方法对缺失值进行更加精确的预测是今后进一步研究的重点方向。【结论】结合了MIC和迁移学习的TLM方法能提升数据预测的准确率,可为SDGs相关领域工作者在处理数据缺失问题时提供重要的参考价值。
文摘目的分析2010-2017年我国31个省市道路交通事故发生数和死伤情况,预测到2020年各省市实现可持续发展目标(sustainable development goals,SDGs)中设定的死伤人数较2015年减半的情况。方法分析31个省市道路交通事故现状,利用趋势外推法预测2020年能否实现SDGs目标。结果2010-2017年广东、江苏、浙江、山东的交通事故发生数、死伤数呈下降趋势但仍居前列。2017年死亡率前两位是北京和贵州,受伤率贵州和天津居高。预测显示,到2020年全国道路交通事故的死亡和受伤人数分别增加了17%和1.3%,难以实现减半的目标;预计2020年湖南的死伤人数分别降至2015年的50.7%和65.3%,能够达到目标;上海的受伤人数降低83.3%,但死亡人数只降低34.5%,与目标有差距;其余省市难以达到目标。预计到2020年11个省市的死亡人数将增加,其中北京、吉林、江西、湖北和贵州增长最为明显;有8个省市的受伤人数呈增长趋势,其中吉林、江西、湖北和贵州四省增幅明显。结论除湖南省外,全国和各省市均难以达到SDG s目标,可以根据预测结果和实际现状制定有针对性的科学防控策略。湖南和上海在交通事故防控方面的措施有效,值得借鉴。