As the global economic crisis deepens,people have a clearer understanding of risks.Since internal control and corporate governance can better control risks in the development of a company,how to properly handle the re...As the global economic crisis deepens,people have a clearer understanding of risks.Since internal control and corporate governance can better control risks in the development of a company,how to properly handle the relationship between internal control and corporate governance has become the focus of research.There are mainly three views on the relationship between the two among scholars’researches including environmental theory,basic theory and chimerism theory.In the normal operation of a company,as internal control and corporate governance are closely related,people are more inclined to the point of view of chimerism,claiming that the two affect each other and they are inseparable.展开更多
Museums’building of international communication capacity has become a hot topic in museology.Featuring unique resources of ancient Shu culture,Sanxingdui Museum is of great value in international communication.This p...Museums’building of international communication capacity has become a hot topic in museology.Featuring unique resources of ancient Shu culture,Sanxingdui Museum is of great value in international communication.This paper first reviews previous studies in this area,then analyzes the efforts made by present-day museums in China and abroad to raise their international profile,and later probes into the current status of Sanxingdui Museum’s building of international communication capacity.On this basis,this paper summarizes the countermeasures Sanxingdui Museum took and proposes suggestions regarding the building of its international communication capacity.Sanxingdui Museum should work harder to improve its overseas exhibitions,protect and study its cultural relics,advance the application for World Heritage status,and promote the Sanxingdui brand via multimedia.展开更多
时序抽象作为分层强化学习的重要研究内容,允许分层强化学习智能体在不同的时间尺度上学习策略,可以有效解决深度强化学习难以处理的稀疏奖励问题。如何端到端地学习到优秀的时序抽象策略一直是分层强化学习研究面临的挑战。Option-Crit...时序抽象作为分层强化学习的重要研究内容,允许分层强化学习智能体在不同的时间尺度上学习策略,可以有效解决深度强化学习难以处理的稀疏奖励问题。如何端到端地学习到优秀的时序抽象策略一直是分层强化学习研究面临的挑战。Option-Critic(OC)框架在Option框架的基础上,通过策略梯度理论,可以有效解决此问题。然而,在策略学习过程中,OC框架会出现Option内部策略动作分布变得十分相似的退化问题。该退化问题影响了OC框架的实验性能,导致Option的可解释性变差。为了解决上述问题,引入互信息知识作为内部奖励,并提出基于互信息优化的Option-Critic算法(Option-Critic Algorithm with Mutual Information Optimization,MIOOC)。MIOOC算法结合了近端策略Option-Critic(Proximal Policy Option-Critic,PPOC)算法,可以保证下层策略的多样性。为了验证算法的有效性,把MIOOC算法和几种常见的强化学习方法在连续实验环境中进行对比实验。实验结果表明,MIOOC算法可以加快模型学习速度,实验性能更优,Option内部策略更有区分度。展开更多
文摘As the global economic crisis deepens,people have a clearer understanding of risks.Since internal control and corporate governance can better control risks in the development of a company,how to properly handle the relationship between internal control and corporate governance has become the focus of research.There are mainly three views on the relationship between the two among scholars’researches including environmental theory,basic theory and chimerism theory.In the normal operation of a company,as internal control and corporate governance are closely related,people are more inclined to the point of view of chimerism,claiming that the two affect each other and they are inseparable.
文摘Museums’building of international communication capacity has become a hot topic in museology.Featuring unique resources of ancient Shu culture,Sanxingdui Museum is of great value in international communication.This paper first reviews previous studies in this area,then analyzes the efforts made by present-day museums in China and abroad to raise their international profile,and later probes into the current status of Sanxingdui Museum’s building of international communication capacity.On this basis,this paper summarizes the countermeasures Sanxingdui Museum took and proposes suggestions regarding the building of its international communication capacity.Sanxingdui Museum should work harder to improve its overseas exhibitions,protect and study its cultural relics,advance the application for World Heritage status,and promote the Sanxingdui brand via multimedia.
文摘时序抽象作为分层强化学习的重要研究内容,允许分层强化学习智能体在不同的时间尺度上学习策略,可以有效解决深度强化学习难以处理的稀疏奖励问题。如何端到端地学习到优秀的时序抽象策略一直是分层强化学习研究面临的挑战。Option-Critic(OC)框架在Option框架的基础上,通过策略梯度理论,可以有效解决此问题。然而,在策略学习过程中,OC框架会出现Option内部策略动作分布变得十分相似的退化问题。该退化问题影响了OC框架的实验性能,导致Option的可解释性变差。为了解决上述问题,引入互信息知识作为内部奖励,并提出基于互信息优化的Option-Critic算法(Option-Critic Algorithm with Mutual Information Optimization,MIOOC)。MIOOC算法结合了近端策略Option-Critic(Proximal Policy Option-Critic,PPOC)算法,可以保证下层策略的多样性。为了验证算法的有效性,把MIOOC算法和几种常见的强化学习方法在连续实验环境中进行对比实验。实验结果表明,MIOOC算法可以加快模型学习速度,实验性能更优,Option内部策略更有区分度。
