445 research outputs found

    Semi-linear credibility results

    No full text
    An original paper which suggests a way of thinking for semilinear credibility theory develpment, founded on analysis of the functions of the observable random variables. This line of thought fits perfectly within the framework of the greatest accuracy credibility theory.linear functions, the transformed observations, semi-linear credibility estimators

    Adaptive multiagent system for seismic emergency management

    No full text
    Presently, most multiagent frameworks are typically programmed in Java. Since the JADE platform has been recently ported to .NET, we used it to create an adaptive multiagent system where the knowledge base of the agents is managed using the CLIPS language, also called from .NET. The multiagent system is applied to create seismic risk scenarios, simulations of emergency situations, in which different parties, modeled as adaptive agents, interact and cooperate.adaptive systems, risk management, seisms.

    More general credibility models

    No full text
    summary:This communication gives some extensions of the original Bühlmann model. The paper is devoted to semi-linear credibility, where one examines functions of the random variables representing claim amounts, rather than the claim amounts themselves. The main purpose of semi-linear credibility theory is the estimation of μ0(θ)=E[f0(Xt+1)θ]\mu _0 (\theta ) = E[f_0 (X_{t+1})| \theta ] (the net premium for a contract with risk parameter θ\theta ) by a linear combination of given functions of the observable variables: X=(X1,X2,,Xt)\underline X' = (X_1, X_2, \ldots , X_t). So the estimators mainly considered here are linear combinations of several functions f1,f2,,fnf_1, f_2, \ldots , f_n of the observable random variables. The approximation to μ0(θ)\mu _0 (\theta ) based on prescribed approximating functions f1,f2,,fnf_1, f_2, \ldots , f_n leads to the optimal non-homogeneous linearized estimator for the semi-linear credibility model. Also we discuss the case when taking fp=ff_p = f for all pp to find the optimal function ff. It should be noted that the approximation to μ0(θ)\mu _0 (\theta ) based on a unique optimal approximating function ff is always better than the one in the semi-linear credibility model based on prescribed approximating functions: f1,f2,,fnf_1, f_2, \ldots , f_n. The usefulness of the latter approximation is that it is easy to apply, since it is sufficient to know estimates for the structure parameters appearing in the credibility factors. Therefore we give some unbiased estimators for the structure parameters. For this purpose we embed the contract in a collective of contracts, all providing independent information on the structure distribution. We close this paper by giving the semi-linear hierarchical model used in the applications chapter

    Competition and Antitrust Policy in the Enlarged European Union: A Level Playing Field?

    No full text
    With the central and eastern European countries (CEECs) increasingly included into the international division of labour in the European economic space, we are prompted to ask whether this integration operates on a level playing field with respect to competition policy. In fact, our analysis reveals that effectiveness of implementation of competition law and policy and intensity of competition are lower in the CEECs. We find no reason to believe that the new eastern EU members struggle with the recent reforms of competition policy in the EU, nor do we see the necessity for policy action to spur effective implementation. Copyright (c) 2009 The Author(s). Journal compilation (c) 2009 Blackwell Publishing Ltd.

    STRATEGY MANAGEMENT IN A MULTI-AGENT SYSTEM USING NEURAL NETWORKS FOR INDUCTIVE AND EXPERIENCE-BASED LEARNING

    No full text
    Intelligent agents and multi-agent systems prove to be a promising paradigm for solving problems in a distributed, cooperative way. Neural networks are a classical solution for ensuring the learning ability of agents. In this paper, we analyse a multi-agent system where agents use different training algorithms and different topologies for their neural networks, which they use to solve classification and regression problems provided by a user. Out of the three training algorithms under investigation, Backpropagation, Quickprop and Rprop, the first demonstrates inferior performance to the other two when considered in isolation. However, by optimizing the strategy of accepting or rejecting tasks, Backpropagation agents succeed in outperforming the other types of agents in terms of the total utility gained. This strategy is learned also with a neural network, by processing the results of past experiences. Therefore, we show a way in which agents can use neural network models for both external purposes and internal ones.agents, learning, neural networks, strategy management multi-agent system.

    The problem of determining estimators for different structural parameters in the case of credibility results for weighted contracts

    No full text
    summary:This paper presents and analyzes the estimators of the structural parameters, in the Bühlmann-Straub model, involving complicated mathematical properties of conditional expectations and of conditional covariances. So to enable to use the better linear credibility results obtained in this model, we will provide useful estimators for the structure parameters. From the practical point of view it is stated the attractive property of unbiasedness for these estimators

    Competition Law and Human Rights:Striking a Balance Between Business Freedom and Regulatory Intervention

    No full text
    Presented at The Limits of Competition Law, Santorini, Greece (jointly organised by UCL Faculty of Laws and IMEDIPA)</p

    Parikh Matrices and Istrail Morphism

    No full text
    A word w is a sequence of symbols. A scattered subword or simply a subword u of the word w is a subsequence of w. Parikh matrix M(w) is an ingenius tool introduced by Mateescu et al (2001) to count certain subwords in a word w. Various properties of Parikh matrices have been established. Two words u and v are said to be M-ambiguous or amiable if their Parikh matrices M(u) and M(v) are the same. On the other hand a morphism f is a mapping on words w whose images f(w) are also words with the property that, f(uv)=f(u)f(v) for given words u and v. Istrail morphism (Istrail, 1977) is a specific kind of morphism on a set {a,b,c} of three symbols. Using this morphism, M-ambiguity or amiability of words based on Parikh matrices is investigated by Atanasiu (2010). Parikh matrices of words that involve certain ratio-property are investigated by Subramanian et al (2009). Here we consider this kind of ratio-property in the context of Istrail morphism and obtain certain properties of morphic images of words under Istrail morphism. Using these properties, conditions are obtained for product of Parikh matrices of such morphic images under Istrail morphism to commute
    corecore