通过选取并提交专家标注最有信息量的样例,主动学习算法中可以有效地减轻标注大量未标注样例的负担.采样是主动学习算法中一个影响性能的关键因素.当前主流的采样算法往往考虑选取的样例尽可能平分版本空间.但这一方法假定版本空间中的...通过选取并提交专家标注最有信息量的样例,主动学习算法中可以有效地减轻标注大量未标注样例的负担.采样是主动学习算法中一个影响性能的关键因素.当前主流的采样算法往往考虑选取的样例尽可能平分版本空间.但这一方法假定版本空间中的每一假设都具有相同的概率成为目标函数,而这在真实世界问题中不可能满足.分析了平分版本策略的局限性.进而提出一种旨在尽可能最大限度减小版本空间的启发式采样算法MPWPS(the most possibly wrong-predicted sampling),该算法每次采样时选取当前分类器最有可能预测错误的样例,从而淘汰版本空间中多于半数的假设.这种方法使分类器在达到相同的分类正确率时,采样次数比当前主流的针对平分版本空间的主动学习算法采样次数更少.实验表明,在大多数数据集上,当达到相同的目标正确率时,MPWPS方法能够比传统的采样算法采样次数更少.展开更多
文章基于2018年中国家庭追踪调查(China Family Panel Studies,CFPS)数据库,采用可加的半参数分位数模型,研究家庭老龄人口对居民消费的影响。研究发现:家庭的老龄人口越多,家庭平均消费支出越少,家庭老龄人口指标对家庭平均消费支出起...文章基于2018年中国家庭追踪调查(China Family Panel Studies,CFPS)数据库,采用可加的半参数分位数模型,研究家庭老龄人口对居民消费的影响。研究发现:家庭的老龄人口越多,家庭平均消费支出越少,家庭老龄人口指标对家庭平均消费支出起到消极阻碍作用。而老年家庭、年轻家庭、混合家庭在各项消费支出中具有显著差异。从消费结构的横向来比,家庭支出最多的都是饮食支出。老年家庭在医疗保健方面的支出占比高于年轻家庭和混合家庭。因此,政府应该关注老龄人的消费需求,在相应政策上予以倾斜照顾,创造更适宜的养老环境。展开更多
The current work is an extension of the nonlocal elasticity theory to fractional order thermo-elasticity in semiconducting nanostructure medium with voids.The analysis is made on the reflection phenomena in context of...The current work is an extension of the nonlocal elasticity theory to fractional order thermo-elasticity in semiconducting nanostructure medium with voids.The analysis is made on the reflection phenomena in context of three-phase-lag thermo-elastic model.It is observed that,four-coupled longitudinal waves and an independent shear vertical wave exist in the medium which is dispersive in nature.It is seen that longitudinal waves are damped,and shear wave is un-damped when angular frequency is less than the cut-off frequency.The voids,thermal and non-local parameter affect the dilatational waves whereas shear wave is only depending upon non-local parameter.It is found that reflection coefficients are affected by nonlocal and fractional order parameters.Reflection coefficients are calculated analytically and computed numerically for a material,silicon and discussed graphically in details.The results for local(classical)theory are obtained as a special case.The study may be useful in semiconductor nanostructure,geology and seismology in addition to semiconductor nanostructure devices.展开更多
文摘通过选取并提交专家标注最有信息量的样例,主动学习算法中可以有效地减轻标注大量未标注样例的负担.采样是主动学习算法中一个影响性能的关键因素.当前主流的采样算法往往考虑选取的样例尽可能平分版本空间.但这一方法假定版本空间中的每一假设都具有相同的概率成为目标函数,而这在真实世界问题中不可能满足.分析了平分版本策略的局限性.进而提出一种旨在尽可能最大限度减小版本空间的启发式采样算法MPWPS(the most possibly wrong-predicted sampling),该算法每次采样时选取当前分类器最有可能预测错误的样例,从而淘汰版本空间中多于半数的假设.这种方法使分类器在达到相同的分类正确率时,采样次数比当前主流的针对平分版本空间的主动学习算法采样次数更少.实验表明,在大多数数据集上,当达到相同的目标正确率时,MPWPS方法能够比传统的采样算法采样次数更少.
文摘文章基于2018年中国家庭追踪调查(China Family Panel Studies,CFPS)数据库,采用可加的半参数分位数模型,研究家庭老龄人口对居民消费的影响。研究发现:家庭的老龄人口越多,家庭平均消费支出越少,家庭老龄人口指标对家庭平均消费支出起到消极阻碍作用。而老年家庭、年轻家庭、混合家庭在各项消费支出中具有显著差异。从消费结构的横向来比,家庭支出最多的都是饮食支出。老年家庭在医疗保健方面的支出占比高于年轻家庭和混合家庭。因此,政府应该关注老龄人的消费需求,在相应政策上予以倾斜照顾,创造更适宜的养老环境。
文摘The current work is an extension of the nonlocal elasticity theory to fractional order thermo-elasticity in semiconducting nanostructure medium with voids.The analysis is made on the reflection phenomena in context of three-phase-lag thermo-elastic model.It is observed that,four-coupled longitudinal waves and an independent shear vertical wave exist in the medium which is dispersive in nature.It is seen that longitudinal waves are damped,and shear wave is un-damped when angular frequency is less than the cut-off frequency.The voids,thermal and non-local parameter affect the dilatational waves whereas shear wave is only depending upon non-local parameter.It is found that reflection coefficients are affected by nonlocal and fractional order parameters.Reflection coefficients are calculated analytically and computed numerically for a material,silicon and discussed graphically in details.The results for local(classical)theory are obtained as a special case.The study may be useful in semiconductor nanostructure,geology and seismology in addition to semiconductor nanostructure devices.
