中国汽车保险的最优索赔策略

在大多数国家,汽车保险中使用的无索赔奖励系统(BMS)只考虑了索赔次数,而我国2006年7月开始实施的A、B、C三个车险条款中的C条款是与索赔额有关的BMS,所以需要对现有的最优索赔策略模型进行推广。本文应用马尔科夫最优化原理推广了汽车保险的最优索赔策略模型,并对我国现行的三个车险条款的最优索赔策略问题进行了实证研究。     

神经网络在波谱中的应用：拉曼光谱数据处理

本文深入研究了神经网络在波谱中的应用。考察了神经网络应用于拉曼光谱数据处理的各种条件及性能，包括学习速率常数（ｌｅａｒｎｉｎｇｒａｔｅｐａｒａｍｅｍｒ）β与传递函数（ｔｒａｎｓｆｅｒｆｕｎｃｔｉｏｎ）ｆ（ｘ）的影响，发现采用非线性双曲线正切ｓｉｇｍｏｉｄ函数和较强的学习率β＝０．５时，可获得较佳的神经网络。对原始谱图及处理后谱图进行比较，证实了神经网络能够提高谱线分辨力，修正波谱微小变形及获得良好的信噪比，说明神经网络可在拉曼光谱中获得很好的应用。     

振动工具传至手臂系统的能量

振动白指病(VWF)是手持振动工具操作工人的职业病,其发病率决定于从工具传入手中的总能量水平。本文介绍了传入手中的能量的计算方法,并结合对操作工人的实测结果加以比较。结论指出,在6.3～1000Hz范围以内,由风镐手柄传入手中的单位时间的功为1.57×10~2J·S~(-1),而由风镐凿子传入手中的单位时间的功则更多。     

Bounds for the optimal critical claim size of a bonus system

The optimal critical claim size of a bonus system determines whether to file a claim with the insurance company after having an accident. The aim of this paper is to demonstrate, within the framework of a simple model, how bounds for the optimal critical claim size can be constructed when only incomplete information on the claim amount distribution is available.     

Optimal multi-period insurance contracts

In multi-period insurance contracts (such as automobile insurance contracts), unlike single-period ones, the premiums that the insured must pay increase whenever he files a claim. Hence, the buyer faces a problem that is absent in one-period models, namely: he must determine for which damages he should file a claim and for which he should not.The optimal claims policy of the buyer is presented for a large class of insurance contracts. It is shown that the buyer will file a claim only if it is larger than some critical value. Based on this it is shown that the buyer prefers a contract that provides full coverage above a deductible for damages that exceed his critical value. In this case the optimal contract is not unique since the buyer is indifferent to the form of the contract for damages below his critical value. It is shown, however, that as in one-period models (Arrow (1963, 1974)) there exists an optimal contract that provides full coverage above a deductible. In multi-period setting, however, the buyer will file a claim only if the damage is sufficiently higher than the deductible.It is also shown that the buyer prefers a strictly positive deductible. Unlike the one-period case (Mossin (1968)), this result holds true even if the premium rates equal the expected payments.     

CONSTANT VOLUMES PER OXYGEN OF THE EARTH''''S LAYERS

Through the calculation of the VPOs (volume per oxygen) of 1698 rocks and minerals, the authors reach the following conclusion: the mean VPO of glassy extrusive rocks is 22.5. The mean VPO of semicrystalline extrusive rocks is 21.5. Crystalline extrusive and all intrusive rooks from granite to gabbro have a mean VPO of 20.6. According to the models of the earth's layers, the VPO of the upper crust is 20.8. If the lower crust is composed of gabbro, the VPO is about the same as that of the upper crust. If it consists of amphibolite or dioritic garnet granulite, the VPO is 18.8. Based on the data of mantle inclusions and high pressure and high temperature experiments, the calculated VPO of the upper mantle is 17.7. All these are constants in each layer and correspond to the packing concentrations in the packing of equal spheres.     

OPTIONS FOR THE REGULATION OF MEDITERRANEAN DEMERSAL FISHERIES

Research and management actions are reviewed with respect to demersal fisheries of the Mediterranean since the Second World War, as reflected in the activities of the General Fisheries Council for the Mediterranean, (GFCM). The scientific background to the priority concern expressed for minimum size limits in the 1960's and 1970's is discussed, and in particular, the mesh selectivity experiments that formed the basis for yield per recruit calculations, with respect to the trawl fishery. More recent considerations, changing our perception of the appropriateness of size at first capture of demersal fish as a management tool in trawl fisheries, are reviewed. It is concluded that for multispecies fisheries where the first priority for fishing effort control is not respected, size limits based on size at maturity, rather than yield per recruit criteria, are more feasible, but that changes in mesh size need to take into account subsequent changes in equity between inshore and offshore fleets, and changes in species composition and areas of distribution during the life history. They also need to consider the high landed value of small fish in many Mediterranean fisheries. Alternative, or supplementary, measures to mesh size regulation that affect capture of small fish are also reviewed, including seasonal closures, closed areas, bans on trawling inshore, and regulations on minimum size at sale. A range of problems to be considered prior to deciding on an increase in mesh size are reviewed, including changes in total effort exerted, changes in increases in fishing power (and especially the impacts on the spawning stock), changes in discard rate, “meshing” of small fish, and indirect mortality during fishing. A strategy for introducing new mesh sizes is suggested, with emphasis, where possible, on the experimental approach, and on supplementary measures to control fishing effort. The paper concludes by considering an alternative paradigm to minimum size regulation for demersal fisheries management; namely, the exploitation of juvenile fish, with provision for escapement of a small proportion of large, mature fish offshore, for which exploitation rate declines and remains low. It is suggested that this strategy may be, de facto, the one prevailing in the small mesh size inshore trawl fishery prior to development of offshore fisheries. The implications of this possibility have to be considered seriously if high effort levels are to be maintained while effective size limits are raised.     

用Poisson-Tweedie模型拟合索赔次数

本文主要考虑用Poisson Tweedie分布族来近似实际索赔次数的分布。在这里,我们把使X2拟合优度统计量的值的大小作为评估的标准,最后用来自B櫣hlmann的一组数据对此方法进行了验证。     

A stochastic calculus model of continuous trading: Complete markets

A paper by the same authors in the 1981 volume of Stochastic Processes and Their Applications presented a general model, based on martingales and stochastic integrals, for the economic problem of investing in a portfolio of securities. In particular, and using the terminology developed therein, that paper stated that every integrable contingent claim is attainable (i.e., the model is complete) if and only if every martingale can be represented as a stochastic integral with respect to the discounted price process. This paper provides a detailed proof of that result as well as the following: The model is complete if and only if there exists a unique martingale measure.