摘要:

ABSTRACTBox and Jenkins [3] have specified a procedure for the development of a “transfer function model,” a model which expresses the interrelationships between two time series. Involving the iterative repetition of identification, estimation, and checking stages, this procedure is comparable to their procedure for the development of an autoregressive‐intergrated‐moving average model for a single time series. The transfer function model development procedure has not been widely applied due to the absence of explanations understandable to the non‐statistician and reasonably priced computer algorithms.The intent of this paper is to provide the non‐statistician with an explanation of the identification stage of the Box‐Jenkins transfer function model development procedure. An extensive class of possible models is logically developed. The identification stage is specified step‐by‐step and is illustrated by three pairs of generated time series analyzed using computer algorithms written by the author. The paper presumes a general knowledge of the Box‐Jenkins identification procedure for s

ISSN:0011-7315    

DOI:10.1111/j.1540-5915.1977.tb01090.x

出版商:Blackwell Publishing Ltd    

年代:1977

数据来源: WILEY

摘要:

ABSTRACTThis paper explores the effect which the choice of critical values in sequential decisions has on the overall distribution of payoffs. It demonstrates in the context of a simple two‐stage sequential decision model that the mean and variance of the overall distribution can be highly sensitive to the choice of critical values. Finally, it considers the problem of selecting optimal critical values when a standard meanvariance approach is use

ISSN:0011-7315    

DOI:10.1111/j.1540-5915.1977.tb01091.x

出版商:Blackwell Publishing Ltd    

年代:1977

数据来源: WILEY

摘要:

ABSTRACTThis article describes the application of a goal programming model to set safety stocks in a fixed‐interval, variable order‐quantity inventory system. The firm is constrained by limited storage space and capital in the determination of inventory levels.The costs resulting from the use of the proposed model are contrasted with costs resulting from an alternative method based solely upon a forecasting model. The costs considered are holding costs, stock‐out costs, and the costs of acquiring excess resources.Excess resources are required when safety‐stock levels, determined by the forecasting model, would utilize more than the available capital and space. The proposed model eliminates the need for acquiring excess resources at the risk of increasing the frequency and size of stock outs. A trade‐off occurs between the potential for greater stock‐out costs versus the costs of acquiring excess resources‐borrowing additional funds or renting additional space.The goal programming model was used to set safety stocks for 15 product groups over a 46 month evaluation period for a multiproduct firm. Significant reductions in costs can be realized if the proposed

ISSN:0011-7315    

DOI:10.1111/j.1540-5915.1977.tb01092.x

出版商:Blackwell Publishing Ltd    

年代:1977

数据来源: WILEY

摘要:

ABSTRACTUsing a Conversational Modelling Language (CML), a special computer language has been implemented to allow the specification of mathematical programming models and their parameterization. The language has been designed to be “natural” in the mathematical programming context and to include interfaces with computer routines, such as MPSX, for optimization. Features implemented include equation writing, preparation of computer input, automatic checking and removal of redundancies, and the ability to display solutions given sufficient specification of decision variables. Examples of the use of the language in building interactive decision support systems are gi

ISSN:0011-7315    

DOI:10.1111/j.1540-5915.1977.tb01093.x

出版商:Blackwell Publishing Ltd    

年代:1977

数据来源: WILEY

摘要:

ABSTRACTA major difficulty in practicing constrained nonlinear optimization is that different optimization algorithms are effective for different classes of problems. The search algorithms tested were selected because they seemingly had the capability of solving ill‐defined problems. The emphasis of this research was on problems that are realistic and do not necessarily satisfy the classical assumptions of convexity, continuity, unimodality, etc. All algorithms were tested on a variety of problems. A problem was described by three types of attributes: the environmental attributes, the objective function attributes, and the constraint set attributes. The results of the study provide selective logic for choosing an effective search algorith

ISSN:0011-7315    

DOI:10.1111/j.1540-5915.1977.tb01094.x

出版商:Blackwell Publishing Ltd    

年代:1977

数据来源: WILEY

摘要:

ABSTRACTThis paper reports an exploratory study of the development, application, and evaluation of a semi‐Markov model which forecasts personnel mobility rates within a major employer. The accuracy of the model is investigated by comparing the results generated by it to the results from a simple Markov model and to the observed behavior within the organizatio

ISSN:0011-7315    

DOI:10.1111/j.1540-5915.1977.tb01095.x

出版商:Blackwell Publishing Ltd    

年代:1977

数据来源: WILEY

摘要:

ABSTRACTThe purpose of this article is to illustrate the use of multiple discriminant analysis in those cases in which all discriminating variables are qualitative. The authors will show that an appropriate qualitative variable discriminant analysis model is a reformulation of the Bayesian decision theoretic model. Rules for the use of multiple discriminant analysis also are suggested for the cases in which some variables are qualitative and some are measurable on at least an interval scale. The qualitative variable discriminant analysis model is illustrated as an appropriate device for selecting a product version that will minimize a manufacturer's risk of product introduction.

ISSN:0011-7315    

DOI:10.1111/j.1540-5915.1977.tb01096.x

出版商:Blackwell Publishing Ltd    

年代:1977

数据来源: WILEY

摘要:

ABSTRACTThe criteria upon which R&D experiments are evaluated are subjective; therefore in practice it is extremely difficult to obtain a series of weights so that the trade‐offs between the various criterion can be analyzed. The problem is compounded when the experiments represent basic research and no prior history is available for utility estimation. This paper presents a graph‐theoretic interpretation of the selection process based only on the rank‐order of the alternatives thus eliminating the need for utility estimation. Index numbers are then obtained as the solution of an integer program. An application and the theoretical foundations of the procedure are disc

ISSN:0011-7315    

DOI:10.1111/j.1540-5915.1977.tb01097.x

出版商:Blackwell Publishing Ltd    

年代:1977

数据来源: WILEY

摘要:

ABSTRACTIn this paper the authors discuss the curriculum evaluation process. The use of student inputs to this process is supported. A demonstration is given of a methodology for analyzing student attitudes towards areas of study in the curriculum, and their perceptions of the relations between those areas of study and the job market.

ISSN:0011-7315    

DOI:10.1111/j.1540-5915.1977.tb01098.x

出版商:Blackwell Publishing Ltd    

年代:1977

数据来源: WILEY