Effectiveness of Q-learning as a tool for calibrating agent-based supply network models

This paper examines effectiveness of Q-learning as a tool for specifying agent attributes and behaviours in agent-based supply network models. Agent-based modelling (ABM) has been increasingly employed to study supply chain and supply network problems. A challenging task in building agent-based supply network models is to properly specify agent attributes and behaviours. Machine learning techniques, such as Q-learning, can be a useful tool for this purpose. Q-learning is a reinforcement learning technique that has been shown to be an effective adaptation and searching mechanism in distributed settings. In this study, Q-learning is employed by supply network agents to search for ‘optimal’ values for a parameter in their operating policies simultaneously and independently. Methods are designed to identify the ‘optimal’ parameter values against which effectiveness of the learning is evaluated. Robustness of the learning's effectiveness is also examined through consideration of different model settings and scenarios. Results show that Q-learning is very effective in finding the ‘optimal’ parameter values in all model settings and scenarios considered.     

A duality theory of payment systems

We model the Central Bank’s management of intraday liquidity in modern real-time gross settlement systems as a linear programming problem parameterized by different intraday monetary policies, such as reserve requirements, net debit caps and Lombard loans. We then use duality theory to determine the shadow-prices of constraints of each bank. These shadow-prices can be used by the Central Bank to set personalized intraday monetary policies in order to reduce idleness of money and to give a microfoundation of the too-big-to-fail policy.     

Determinants of the method of payment in mergers and acquisitions

We empirically investigate the determinants of the payment form in mergers and acquisitions and introduce new variables on the target and acquirer investment characteristics to evaluate whether the concerns of target and acquirer shareholders are taken into account. Our sample encompasses mergers between publicly listed US companies from 1985 to 2004. Similarly we also consider the determinants of announcement returns using the same set of variables. We establish the relevance of a previously unreported variable for the determination of the payment form, the correlation of returns between target and acquirer, besides the more established determinants hostile takeovers, and defence mechanisms; weak evidence is found for the significance of budget constraints and no evidence for asymmetric information or tax considerations being a relevant factor. We do not find that announcement returns are explained by the variables considered.     

Itchy feet vs cool heads: Flow of funds in an agent-based financial market

Investors tend to move funds when they are unhappy with their current portfolio managers׳ performance. We study the effect of the size of this flow of funds in an agent-based model of the financial market. The model combines the discrete choice approach from agent-based modelling, where all capital is mobile, with the evolutionary finance framework where all growth is endogenous. Our results show that, if investors exhibit recency bias in evaluating portfolio managers׳ performance, even a small amount of freely flowing capital has a huge impact on the market dynamics and the survival of noise traders. We also find that investors׳ intensity of choice is a driving force for excess volatility and extreme price movements when the size of the flow of funds is large.     

An agent-based model of the observed distribution of wealth in the United States

Pareto cautiously asserted that the wealth and income distributions which bear his name are universal, basing his argument on observations of this distribution in many different types of economies. In this paper, we present an agent based model (and a scalable approximation of it) in a closely related spirit. The central feature of this model is that wealth enables an individual to secure more wealth. Specifically, the important and novel feature of this model is its ability to simultaneously produce both the Pareto distribution observed in empirical data for the top 10% of the population and the exponential distribution observed for the lower 90%. We show that the model produces these distributions of wealth when initialized with an equitable distribution. Then, using historical data, we initialize the model with US wealth shares in 1988 and show that the model tracks wealth share changes from 1988 to 2012. Simulations to 2088 project that the top 0.01% of the population will possess more than 70% of the total wealth in the economy.     

股票市场流动性溢价的实证研究

资产流动性的高低是否影响资产的价格一直是资本市场理论研究的热点问题，也是投资者决策的重要理论依据之一。本文根据股票市场流动性溢价原理，选取换手率与Amivest流动比率作为股票流动性的衡量指标，采用LR两阶段截面回归方法与似无关回归（SUR）估计法，对上海股票市场的股票流动性与预期收益率的关系进行了实证研究。结果表明，上海股票市场存在显著的流动性溢价，换手率低或Amivest流动比率低，流动性较差的资产具有较高的预期收益。研究同时发现，上海股票市场具有很强的规模效应和价值效应。     

An agent-based model of innovation diffusion: network structure and coexistence under different information regimes

The paper analyzes how the structure of social networks affects innovation diffusion and competition under different information regimes. Diffusion is modeled as the result of idiosyncratic adoption thresholds, local network effects and information diffusion (broadcasting and demonstration effect from previous adopters). A high social cohesion decreases the probability of one innovation cornering the market. Nonetheless, with imperfect information, in small-world networks the higher speed of diffusion produced by the low average distance increases this probability. A low social cohesion also increases the probability of falling into traps of under-adoption. However, such probability is significantly lower with imperfect information, because such regime is characterized by higher levels of market concentrations and this reduces the frictions due to the coexistence of non-compatible product innovations.     

流动性的股票特征——基于中国证券市场的经验数据分析

在对流动性内涵认识的基础上,本文提出股票流动性的本质是股票与现金之间相互转化的能力,满足的是投资者正常的交易需求。从流动性的公司规模和成交金额特征、流动性与股价的关系和流动性水平的稳定性三个方面理论探讨和实证检验了流动性的股票特征,研究结果表明：（1）公司规模一定,可实现的成交金额越多,流动性越好;成交金额一定,公司规模越大,流动性越好;（2）股票的价格与股票流动性之间具有显著的正相关关系;（3）流动性水平在一定时期内具有稳定性。     

Concept-based query language approach to enterprise information systems

In enterprise information systems (EISs) it is necessary to model, integrate and compute very diverse data. In advanced EISs the stored data often are based both on structured (e.g. relational) and semi-structured (e.g. XML) data models. In addition, the ad hoc information needs of end-users may require the manipulation of data-oriented (structural), behavioural and deductive aspects of data. Contemporary languages capable of treating this kind of diversity suit only persons with good programming skills. In this paper we present a concept-oriented query language approach to manipulate this diversity so that the programming skill requirements are considerably reduced. In our query language, the features which need technical knowledge are hidden in application-specific concepts and structures. Therefore, users need not be aware of the underlying technology. Application-specific concepts and structures are represented by the modelling primitives of the extended RDOOM (relational deductive object-oriented modelling) which contains primitives for all crucial real world relationships (is-a relationship, part-of relationship, association), XML documents and views. Our query language also supports intensional and extensional–intensional queries, in addition to conventional extensional queries. In its query formulation, the end-user combines available application-specific concepts and structures through shared variables.     

An analytical model for the formation of economic clusters

A simple spatial economy derived from microeconomic foundations is presented to gain insight into the formation of economic clusters. In this model, the formation of economic clusters is a consequence of the competition between economic forces that are consistent with atomistic agents maximizing their utility. An analytic approach is used to obtain the evolution of economic clusters. With this approach, the number of clusters which will grow can be predicted. These results are derived in the traditional one-dimensional geometry and extended to the more realistic two-dimensional landscape.     

Managing security risks for inter-organisational information systems: a multiagent collaborative model

Information sharing across organisations is critical to effectively managing the security risks of inter-organisational information systems. Nevertheless, few previous studies on information systems security have focused on inter-organisational information sharing, and none have studied the sharing of inferred beliefs versus factual observations. In this article, a multiagent collaborative model (MACM) is proposed as a practical solution to assess the risk level of each allied organisation’s information system and support proactive security treatment by sharing beliefs on event probabilities as well as factual observations. In MACM, for each allied organisation’s information system, we design four types of agents: inspection agent, analysis agent, control agent, and communication agent. By sharing soft findings (beliefs) in addition to hard findings (factual observations) among the organisations, each organisation’s analysis agent is capable of dynamically predicting its security risk level using a Bayesian network. A real-world implementation illustrates how our model can be used to manage security risks in distributed information systems and that sharing soft findings leads to lower expected loss from security risks.     

流动性对中国商业银行业绩影响的实证研究

金融危机反映了银行在面对资金压力时持有的流动性资产不足,同时也凸显了银行流动性监管的重要性。为了建立既能够确保金融体系稳定,又能够兼顾银行盈利能力最大化目标的监管标准,有必要对银行流动性资产对业绩的影响进行实证研究,本文利用2003～2010年中国14家上市商业银行数据,检验了流动性资产对于银行业绩的影响,研究发现流动性资产和银行业绩之间存在非线性关系,而且四大国有商业银行和十家股份制商业银行流动性管理行为存在显著差异。     

An agent-based peer-to-peer architecture for semantic discovery of manufacturing services across virtual enterprises

With the development of virtual enterprise (VE) paradigm, the usage of serviceoriented architecture (SOA) is increasingly being considered for facilitating the integration and utilisation of distributed manufacturing resources. However, due to the heterogeneous nature among VEs, the dynamic nature of a VE and the autonomous nature of each VE member, the lack of both sophisticated coordination mechanism in the popular centralised infrastructure and semantic expressivity in the existing SOA standards make the current centralised, syntactic service discovery method undesirable. This motivates the proposed agent-based peer-to-peer (P2P) architecture for semantic discovery of manufacturing services across VEs. Multi-agent technology provides autonomous and flexible problemsolving capabilities in dynamic and adaptive VE environments. Peer-to-peer overlay provides highly scalable coupling across decentralised VEs, each of which exhibiting as a peer composed of multiple agents dealing with manufacturing services. The proposed architecture utilises a novel, efficient, two-stage search strategy – semantic peer discovery and semantic service discovery – to handle the complex searches of manufacturing services across VEs through fast peer filtering. The operation and experimental evaluation of the prototype system are presented to validate the implementation of the proposed approach.     

Classification systems and selection systems: The risks of radical innovation and category spanning

This paper looks at the relations between classification and competition, using the theoretical framework of selection system theory. In a particular competitive arena more than one classification system can be in use. The proportion of classification systems in use by consumers that are also in use by the dominant selectors is a determinant of the risks of category spanning and radical innovation. Furthermore, these risks can be better understood by taking into account whether the selection system has different stages occupied by selectors of different types and by looking at the relations between the selectors of the primary market and the selectors of these selectors.     

An expanded model of the factors affecting the acceptance and effectiveness of electronic human resource management systems

Despite the widespread use of eHR systems, surveys show that there may be a number of problems associated with their design and implementation [CedarCrestone (2007). CedarCrestone 2007–2008 HR systems survey: HR technologies, service delivery approaches, and metrics. Available at www.cedarcrestone.com/research.php. Retrieved July, 2008]. In an effort to overcome these problems we expanded the model of eHR acceptance and effectiveness developed by Stone, Stone-Romero, & Lukaszewski [Stone, D. L., Stone-Romero, E. F., & Lukaszewski, K. (2006). Factors affecting the acceptance and effectiveness of electronic human resource systems. Human Resources Management Review, 16, 229-244]. The expanded model provides a more detailed discussion of the communication processes underlying these systems including the effects of media and message characteristics. In addition, we offer a number of testable hypotheses based on the model that can be used to guide future research on eHR systems.     

An empirical investigation of the roles of prior knowledge and learning activities in technology transfer

This study analyzes 120 university–industry technology transfer projects. A significant positive relationship was found between the learning activities performed by the firm during the development and implementation stages of the technology transfer project and the benefits to that firm from the project. In contrast, prior knowledge of the firm about the existing technology was found to have only a marginal contribution to the project benefits. However, further exploratory analysis based on high and low levels of technical and organizational uncertainty revealed more provocative relationships.     

Learning curves in collaborative planning,forecasting, and replenishment (CPFR) information systems: An empirical analysis from a mobile phone manufacturer

While Collaborative Planning, Forecasting, and Replenishment (CPFR) information systems have been increasingly deployed to improve supply chain operations in a cross section of industries, the extant literature has largely overlooked the learning effects within organizations, thereby resulting in incomplete assessment of their business value. Using an operational-level panel data for nine product lines over 2.5 years, we empirically examine the learning curves in CPFR between Motorola, a mobile phone manufacturer, and one of its U.S.-based national retail partners. We found that the two key components of CPFR, collaborative forecasting (CF) and collaborative replenishment (CR), exhibit distinct learning curves. Forecast accuracy improves immediately following CPFR implementation but the rate of improvement slows over time, whereas inventory levels increase at first and begin decreasing after a period. Further, we found different learning effects in terms of inventory levels when products are later replaced with new form factors. Product replacements have lower inventory levels than their antecedents, at least for low-end products. We discuss important implications for theory and practice at the interface of information systems and operations management.     

旅游景区电子解说系统发展初探

随着旅游的发展,旅游解说系统在景区的作用日益引起人们的注意,而电子旅游解说系统的出现,使看起来简单的解说系统融入了科技的因素。文章分析了电子解说系统的现状及与传统解说系统的区别,在此基础上提出了策划的原则及要素,为景区发展电子解说系统提供了借鉴。     

Incorporating sustainability indicators into a computable general equilibrium model of the scottish economy

In recent years, the notion of sustainable development has begun to figure prominently in the regional, as well as the national, policy concerns of many industrialized countries. Indicators have typically been used to monitor changes in economic, environmental and social variables to show whether economic development is on a sustainable path. In this paper we endogenize individual and composite environmental indicators within an appropriately specified computable general equilibrium modelling framework for Scotland. In principle, at least, this represents a very powerful modelling tool that can inform the policy making process by identifying the impact of any exogenous policy change on the key endogenous environmental and economic indicators. It can also identify the effects of any binding environmental targets on economic activity.