The number of population founders and the size of the population are two important variables in determining how much gene diversity can be retained in a population. A model is developed to determine the most cost-effective balance of those two variables, based on comparing the marginal costs of increasing the number of founders vs. increasing the population size. Marginal costs, in this case, are the costs of increasing the number of founders or the population size by one animal. For a goal of retaining 90% gene diversity for 10-15 generations (approximately 100 years), the current recommendation of 20 effective founders is very close to the least-cost solution when the ratio of the marginal costs is equal to one. However, when the ratio is greater or less than one that number can change considerably.

Various plausible power architectures used in COS and EEEs are discussed in this paper. Each architecture, including currently used architectures. has its advantages and disadvantages. S i n a architectures are application-spedc, it is emphvized that none or the architectures discussed is considered to be optimum. However, distributed architectures have several advantages over the bulk arclutectures used today. IMPACT ON POWERexpected to decrease. However, many more services are expected to be inccrporated in future networks. The net result of this will be an overall increase in the amount of power required to support the network. As the network architecture is expected to be distributed, the network will have remote modules possibly containing multiple microprocessors and memory in scattered through out the network. In addition. these modules may be placed in locations such as the tops of poles and mountains, underground. the curbside, etc.For such applications high power densities coupled with high efficiencies and high reliability under all environment and load conditions will be essential. POWER SYSTEMS ARCHITECTURESThe power needs of future networks will create demands that cannot be met with current power architectures; therefore, the present power architecture should be re-examined. The major factors influencing the choice of power architectures are as follows:Voltage and energy requirements of the load. Higher levels of integration and the implementation of sub-micron technology integrated circuits within the telecommunication load demand a low-voltage, high-quality power system to be reliably supplied to the load.b. The downtime requirements of the load and the attendant downtime requirements of the power c. The overall life of the power system d. The overall efficiency of the power system and the elimination of multiple conversions (ac-to-dc at some voltage level and then again to dc at anothcr level).gracefully.a.supply.e. The flexibility of the power system to grow ne power for the network wa vary according to the type of service and architecture adooted to imolement it. It must be noted that with f. The utility ratio (the ampacity of the power Plant that is in actual U% over the installed ampacity) of the power plant.

No abstract

Recently, Armstrong and Mitroff have joined an important debate over the methods of science. We are afraid that because of the tongue-in-cheek fashion in which it was done that two important issues were intertwined and may not be fully appreciated. The issues involved the practice of science versus the methods and theory versus data.

In operational applications, forecasts .are normally adjusted by applying safety factors to allow for asymmetry in the underlying loss function. This paper considers an extension of the often used linear loss function to situations in which an error can also result in a fixed sum loss.The need for forecasts as inputs to planning and decision models is widely recognized. Forecasts range from subjective, qualitative predictions of the long range future to short range operational projections which are quantitative and highly structured. In the latter case historical statistics normally are available and form the basis for the projection as in sales forecasting for production and inventory control, cash flow forecasts, and forecasts of labor turnover. Analytic procedures include classical time series analysis, regression models, exponential smoothing, and other adaptive procedures, as well as subjective methods such as the use of executive or sales force opinions.It is generally accepted that forecasts can never be completely accurate. To the extent that the loss functions associated with forecast errors are explicitly considered, the usual assumption is that the loss function is symmetric and often that it is quadratic.' This leads to an emphasis on unbiased predictors. Alternatively, maximum likelihood procedures may be used, implying that a forecast should have a high probability of being (approximately) correct. In practice, the users of forecasts are seldom interested in either an unbiased or a most likely forecast. Instead, a correction usually is added to or subtracted from the forecast to implicitly allow for an intuitively felt asymmetry in the costs of positive and negative forecast errors. As planning and decision models have become more structured, it is frequently possible to objectively include these adjustments or safety factors in the model by formally introducing some measure of forecast uncertainty. Therefore, more and more forecasts are being given as probability distributions rather than as point predictions [4], [ 5 ] , [ 6 ] .In some cases, planning and decision models can directly incorporate these probability distributions by using payoff tables or decision trees. In other cases, the forecast variable is quantitative, and the asymmetric loss function can be made an integral part of the statistical analysis [l]. Since a substantial forecasting methodology already exists based on symmetric loss functions, it is far more practical to consider ways of adjusting these forecasts through correction For example see [7, p. 161. 425