Predictive models and generative complexity

The causal states of computational mechanics define the minimal sufficient memory for a given discrete stationary stochastic process.Their entropy is an important complexity measure called statistical complexity(or true measure complexity).They induce the e-machine,which is a hidden Markov model(HMM) generating the process.But it is not the minimal one,although generative HMMs also have a natural predictive interpretation.This paper gives a mathematical proof of the idea that the e-machine is the minimal HMM with an additional(partial) determinism condition.Minimal internal state entropy of a generative HMM is in analogy to statistical complexity called generative complexity.This paper also shows that generative complexity depends on the process in a nice way.It is,as a function of the process,lower semi-continuous(w.r.t.weak-* topology),concave,and behaves nice under ergodic decomposition of the process.     

分裂化恐怖组织袭击策略的Moran过程随机演化模型

国际恐怖组织出现分裂化的演变趋势及多样化的行为诱因.本文构建了分裂化恐怖组织袭击策略的Moran过程演化模型，得出了分别在经济动因、意识形态和外界随机干扰因素主导下不袭击策略达到演化稳定的条件.研究表明，恐怖组织的策略演化主要取决于成本收益因素、恐怖组织的分裂化程度和政府打击恐怖主义力度.对于以经济利益为主导动机的恐怖组织，不袭击策略成为恐怖分子的演化稳定策略需要满足一定的收益条件.对于"独狼"类型的恐怖分子，所在国政府单纯从经济利益角度进行防范无法实现有效遏制.对于以意识形态为主导动机的恐怖组织，袭击策略始终是演化稳定策略.当外界随机干扰因素占据主导地位时，小规模的恐怖组织，以及恐怖组织分裂数目增大，反而使得袭击行动更加频繁.     

云计算环境中柔性易扩展的信任协商机制研究

提出了一种云计算环境中柔性易扩展的信任协商机制-基于相似信任度的信任协商机制,在该信任协商机制中,直接信任度计算过程中加入了为用户节点提供服务的服务时刻和服务成本等信任因素,间接信任度计算过程中加入了参与评价的云服务节点个数、评价用户节点的服务总量、评价用户节点所给评价的相似度等信任因素,并且详细讨论了这些信任因素的获取和计算方法.     

基于绩效反馈机制的组织能力重构研究

针对学界缺乏对能力陷阱识别问题考察的现状,本研究提出组织通过绩效反馈机制识别能力陷阱并进行跨越的动态模型,构建组织行为理论和能力重构理论间的联系,为学界理解能力重构的微观基础提供思路.相比于"组织竞争优势来源于前瞻式的战略预见"的主导逻辑,本文提出更符合管理者有限理性假说的能力重构思路:管理者通过后顾式、问题导向的组织搜索行为实现能力重构.其逻辑在于:"目标驱动的搜索决策"可被视为竞争优势的微观来源——合理的目标驱动搜索行为可以使组织因结构化惯例实现较为彻底的租金占用,这有助于管理者明晰应何时通过路径变异进行新的租金创造,以克服伴随组织惯例的能力锁定陷阱.     

基于信任和不信任关系的实值受限玻尔兹曼机推荐算法

在社交网络与电子商务快速融合的背景下,将基于信任关系的推荐技术应用于电子商务领域实现个性化推荐已得到广泛研究.现有推荐算法鲜有考虑用户间不信任效应,导致社交信任度量过于保守,较大地影响了推荐系统准确性.针对现有推荐算法忽视不信任关系导致的非对称效应缺陷,本文提出一种结合信任和不信任的实值受限玻尔兹曼机推荐算法(TDA-RBM),首先建立个人受限玻尔兹曼机,进而运用用户社交行为特征信息分析用户信任与不信任关系并进行度量,在此基础上构造信任-不信任监督机制并用于TDA-RBM方法的优化,同时对该方法的有效性进行分析.通过Epinions数据进行的对比实验表明了TDA-RBM方法的有效性以及不信任关系的引入能有效提高推荐准确性.     

A Holistic Approach for Designing Human-Centric Trust Systems

Online trust systems, albeit existing as technical implementations, aim to translate the role that trust has in the traditional world onto the virtual platforms. Establishing the inter-dependence between these systems and the human-factor is essential for reducing, but more importantly for understanding the inherent complexity of the open social platforms, and for improving the user-experience and system-performance. This work determines the systemic features of trust and introduces a novel framework of design-properties based on the principles of General Systems Theory. We determine the systemic properties which lack in the current technical solutions, and incorporate social factors into the design-guidelines of trust systems. We introduce four influence-factors and demonstrate that accounting for those factors leads to compliance with the generic system-principles. Finally, we employ case study analysis following the premises of case study research established in the social science literature. By performing semantic analysis of the state of the art on trust systems, we are able to identify two semantic clusters to which trust research belongs, and to choose in a non-random manner the representative case studies for evaluation upon our framework. To our knowledge, this is the first study applying fundamental social science principles from General Systems Theory and case-study research for the purpose of theory-building and evaluation of technical human-centric solutions.     

虚拟企业伙伴选择的信任场模型

虚拟企业是应对市场竞争环境变化的一种柔性的能快速反应的网络化分布式动态组织. 在整个虚拟企业的组建过程中, 如何选择值得信任的合作伙伴是虚拟企业成败的关键. 针对虚拟企业的伙伴选择问题, 本文结合信誉和基于信任的合作机制, 借鉴"场"理论, 提出了信任场的概念, 建立了信任场的理论模型, 并讨论了基于信任场模型的虚拟企业伙伴选择方法. 最后, 通过实例及分析说明了该方法的可用性和合理性.     

A Measure of Learning Model Complexity by VC Dimension

1　IntroductionMeasure of model complexity is important for understanding how difficult a modelingproblem is and for optimizing model performance.Modeling,in all its varied forms,is theprocess of logical induction. A key question regarding induction can be stated as:whatmust one know a priori about an unknown functional dependency in order to estimate itonthe basis of observations?[1 ] . In the fields of engineering and computer science,thisquestion is known as the learning problem. Recently,…     

论“系统科学”与“复杂性科学”之异同

“系统科学”不同于“复杂性科学”,但由于其兴起背景和演化发展过程、研究对象以及指导范式等因素上的相近性,它们实质上又是相互融合的,是易于混淆的两种提法。对它们进行细致的区分是进一步了解系统科学与系统哲学研究领域的必要过程。     

学习型组织的内部控制及其评价

学习型组织被认为是21世纪的最成功的企业模式，但对于如何建立学习型组织并加强其管理控制，目前尚处于探索阶段。从组织学习与内部控制之间的相关性入手，旨在分析学习型组织内部控制的特征、方式和评价方法。     

信任网络、资金流动与区域发展差距

信任被认为是一种非正式制度和重要的社会资本，在正式制度尚未成熟尤其是法律基础薄弱的地区具有重要作用，能够弥补正式制度的缺陷，降低交易成本.利用中国人民银行大额支付系统和中国企业家调查系统的数据，本文首先考察了省际双边信任对地区间资金流动的因果影响.结果表明，来源地对目标地的信任水平会显著促进资金从来源地流入目标地.通过进一步构造基于地理关联、文化关联和信任关联的信任网络，我们发现信任网络的溢出效应对资金流动的促进作用大于来源地对目标地信任的直接影响.此外，两个地区间的双边信任差距越大，会引致更大的资金流动差异，进而扩大区域发展差距.本文从社会资本角度为中国区域发展不平衡现象给予了新的解释，并提出相应的政策建议.     

Industrial organizations as knowledge systems

Industrial organizations are traditionally seen around a nucleus of large-scale mechanized facilities. In today's knowledge society, however, industrial organizations cannot be described sufficiently by mechanistic metaphors. Knowledge has become the decisive factor for corporate growth. It can be created through access to individuals' cumulative experience and wisdom. This paper proposes a model to describe industrial organizations as knowledge systems which communicate and evolve in accordance with their environments. Knowledge which has been previously unknown to an organization emerges out of the interplay of internalization, creation, distribution, and externalization of knowledge. The emergence of knowledge is seen as a self-organizing process, without intervention by the outside environment or an internal designer.     

基于强化学习的战时保障力量调度策略研究

智能化后装保障调度是当前军事领域的研究热点之一,其中复杂多变的战场环境要求战时保障具有良好的自适应性。针对此问题,提出了基于马尔可夫决策过程的强化学习模型,能够主动学习最佳派遣策略,根据历史数据和当前态势预判后续变化。为了考虑不确定事件的影响,在模型求解算法中增加了基于概率统计模型的仿真流程;为了减少随机事件带来的计算复杂性,利用决策后状态变量重新设计了贝尔曼迭代方程;为了解决状态空间的维度灾问题,提出了基于基函数组合的近似函数。仿真实验表明,强化学习能力的引入能够显著提高战时保障调度性能。     

Developing Managers as Healing Agents in Organizations: A Co-Operative Inquiry Approach

This action research examines how a health service manager, with a clinical background, sought to develop his clinical skills in a management context. When asked to undertake an evaluation of an addiction counseling service, he found that there was evidence of systemic ill-health expressed through feelings of powerlessness, hurt, and alienation among the staff of the service. The article describes how the manager used action research, and in particular co-operative inquiry, to become an agent of healing in his own place of work. It illustrates how the co-operative inquiry process can be used to help individuals or groups to overcome feelings of alienation and powerlessness in organizations. By dealing with these feelings through a process of co-operative inquiry the conditions necessary to improve systemic health can be created to help develop a service and meet service objectives It explores the neglected subject of how managers can be agents of healing in organizations.     

Practicing What We Teach in Teaching Systems Practice: The Action–Learning Cycle

Respect for autonomy is a powerful tool for managing complexity. It lets natural, mutually supportive order emerge. In Western culture, though, much order is imposed. This causes conflict, which only increases complexity. This conflictual pattern has an antidote in systems practice: the systemic action–learning cycle. When used reflectively at the level of second-order cybernetics, this cycle embodies respect for autonomy. The UK Open University course "T306: Managing Complexity—A Systems Approach" teaches this action–learning cycle, and uses the cycle in its own teaching. In particular, it uses the cycle in its online conferences, to invite participation and dissolve conflict. This paper shows how.     

A multidimensional perspective of organizational change

This paper discusses some key societal changes which are pressuring organizations to find new ways to use their workforces more creatively so that they can face increasing environmental and internal complexity. It is argued that we should use a multidimensional perspective to promote the multifaceted organizational change needed in today's society, since each existing approach gives a distinctive but partial contribution to the understanding of organizational phenomena. A multidimensional training and intervention technique we have begun to develop to promote organizational change is then presented, and this is compared and contrasted with Critical Systems philosophy, principles, and methods of intervention.     

Reinforcement learning for mobile robot:fromreaction to deliberation

1 .INTRODUCTIONAutonomous mobilerobotics has beena veryi mportant re-searcharea in artificial intelligence and control theory fordecades. Researchers have developed series of methodolo-giesto build mobile robots withself-learning and adaptivecontrol capabilities,rangingfromdeliberative model-basedto purelyreactive paradigms.But nostrongsuccesses havebeen achieved yet . For there are three main difficulties:(1) We cannot programeverything for the robot ,espe-cially in an unknown environmen…     

Managing Complexity—The Path Toward Intelligent Organizations

Coping with complexity is at the heart of management and leadership in the turbulent environments faced by the organizations and societies of our day. The systems approach provides transdisciplinary theories and tools for dealing with this challenge more effectively than efforts merely based on disciplinary insights or pragmatic recipes. In this paper, a concept of "intelligent organizations" is introduced. A framework for the design of intelligent organizations is proposed that links three organizational cybernetic models: the Model for Systemic Control, the Viable System Model, and the Team Syntegrity model.² The proposition is that a combined use of these models, guided by the integrative conceptual framework, enables a more effective response to complex situations than merely pragmatic approaches to "integrative management." Provisional empirical evidence gathered from applied research indicates that the proposed framework is capable of enhancing such superior potential for effective response. As such, it promises to meet the specific needs of the new types of organizations that have already begun to emerge in the knowledge societies.     

Facilitating Trust Building in Networks: A Study from the Water Technology Industry

This paper analyses how action researchers can facilitate trust building processes in inter-firm networks and develops a framework for network development. A longitudinal case study of developing a regional network of water technology SMEs constitutes the empirical base. The paper argues that researchers can directly facilitate processes with a capacity to build two types of trust in different phases of network development, both characteristic-based and process-based trust. The findings indicate that processes to build characteristic-based trust can be facilitated through dialogue processes in temporary groups at network meetings. Processes to build process-based trust are stimulated by practical inter-firm teamwork. Furthermore, there seems to be a mutually reinforcing relationship between these two forms of trust formation, which can be influenced by action researchers. When the level of trust has reached a point of critical mass, new-coming firms seem to jump quickly through characteristic-based trust towards a relatively high level of process-based trust.     

供应链中社会资本、不确定性与机会主义行为关系的实证研究

供应链中普遍存在机会主义行为风险，而关于此类研究成果主要基于交易成本理论而缺乏对供应链中的社会关系的考量。深入分析了社会资本的结构、认知和关系维度的内在关系，给出了社会资本抑制机会主义行为的作用路径。建立了社会资本-不确定性-机会主义行为的结构方程模型，通过采集制造业、石油化工业等不同行业的300份供应链成员有效问卷对模型假设进行检验。研究结果表明：社会资本结构维度中信息交换对认知维度中的文化理解与过程集成均有显著正向相关性，文化理解与关系维度中的信任与合作规范均有显著正向相关性，过程集成与关系维度中的信任有显著负相关性、与合作规范呈现显著正相关性。认知维度中文化理解与供应链成员内部不确定性显著负相关，关系维度中信任与供应链成员行为不确定性显著负相关，而供应链中不确定性显著提升其机会行为发生的可能性。