随着计算机技术的飞速发展,各个领域的数据量呈指数上升,如何处理大体量的数据以保证数据的质量和可用性是机器学习建模过程中不可缺少的一部分。作为建模的首要部分,数据预处理技术的精度直接影响算法的性能。在已有研究的基础上提出...随着计算机技术的飞速发展,各个领域的数据量呈指数上升,如何处理大体量的数据以保证数据的质量和可用性是机器学习建模过程中不可缺少的一部分。作为建模的首要部分,数据预处理技术的精度直接影响算法的性能。在已有研究的基础上提出了一种新颖的数据预处理方法,将数据预处理过程的不同方面集成到一起,构造出基于集成的数据预处理方法,最后在UCI数据库中(UCI数据库是加州大学欧文分校University of California Irvine提出的用于机器学习的数据库)3个经典数据集的基础上进行实证研究,并使用决策树、支持向量机、神经网络这3种机器学习算法来验证集成数据预处理技术的可行性和提高预测性能的有效性。展开更多
The climate has an impact on the urban thermal environment,and the magnitude of the surface urban heat island(SUHI)and urban cool island(UCI)vary across the world’s climatic zones.This literature review investigated:...The climate has an impact on the urban thermal environment,and the magnitude of the surface urban heat island(SUHI)and urban cool island(UCI)vary across the world’s climatic zones.This literature review investigated:1)the variations in the SUHI and UCI intensity under different climatic backgrounds,and 2)the effect of vegetation types,landscape composition,urban configuration,and water bodies on the SUHI.The SUHI had a higher intensity in tropical(Af(tropical rainy climate,Köppen climate classification),Am(tropical monsoon climate),subtropical(Cfa,subtropical humid climate),and humid continental(Dwa,semi-humid and semi-arid monsoon climate)climate zones.The magnitude of the UCI was low compared to the SUHI across the climate zones.The cool and dry Mediterranean(Cfb,temperate marine climate;Csb,temperate mediterranean climate;Cfa)and tropical climate(Af)areas had a higher cooling intensity.For cities with a desert climate(BWh,tropical desert climate),a reverse pattern was found.The difference in the SUHI in the night-time was greater than in the daytime for most cities across the climate zones.The extent of green space cooling was related to city size,the adjacent impervious surface,and the local climate.Additionally,the composition of urban landscape elements was more significant than their configuration for sustaining the urban thermal environment.Finally,we identified future research gaps for possible solutions in the context of sustainable urbanization in different climate zones.展开更多
文摘随着计算机技术的飞速发展,各个领域的数据量呈指数上升,如何处理大体量的数据以保证数据的质量和可用性是机器学习建模过程中不可缺少的一部分。作为建模的首要部分,数据预处理技术的精度直接影响算法的性能。在已有研究的基础上提出了一种新颖的数据预处理方法,将数据预处理过程的不同方面集成到一起,构造出基于集成的数据预处理方法,最后在UCI数据库中(UCI数据库是加州大学欧文分校University of California Irvine提出的用于机器学习的数据库)3个经典数据集的基础上进行实证研究,并使用决策树、支持向量机、神经网络这3种机器学习算法来验证集成数据预处理技术的可行性和提高预测性能的有效性。
基金Under the auspices of the National Natural Science Foundation of China(No.41590841)the National Key Research and Development Program of China(No.2016YFC0503000)the Research Funds of the Chinese Academy of Sciences the Chinese Academy of Sciences(CAS)-the World Academy of Sciences(TWAS)President’s Fellowship。
文摘The climate has an impact on the urban thermal environment,and the magnitude of the surface urban heat island(SUHI)and urban cool island(UCI)vary across the world’s climatic zones.This literature review investigated:1)the variations in the SUHI and UCI intensity under different climatic backgrounds,and 2)the effect of vegetation types,landscape composition,urban configuration,and water bodies on the SUHI.The SUHI had a higher intensity in tropical(Af(tropical rainy climate,Köppen climate classification),Am(tropical monsoon climate),subtropical(Cfa,subtropical humid climate),and humid continental(Dwa,semi-humid and semi-arid monsoon climate)climate zones.The magnitude of the UCI was low compared to the SUHI across the climate zones.The cool and dry Mediterranean(Cfb,temperate marine climate;Csb,temperate mediterranean climate;Cfa)and tropical climate(Af)areas had a higher cooling intensity.For cities with a desert climate(BWh,tropical desert climate),a reverse pattern was found.The difference in the SUHI in the night-time was greater than in the daytime for most cities across the climate zones.The extent of green space cooling was related to city size,the adjacent impervious surface,and the local climate.Additionally,the composition of urban landscape elements was more significant than their configuration for sustaining the urban thermal environment.Finally,we identified future research gaps for possible solutions in the context of sustainable urbanization in different climate zones.