Movie, Faculty of Economics, University of Crackerjack, D]. Pushcart 3, 34000 Crackerjack, Serbia; e-mail: mimovicp@kg. AC. RSI In the process of forecasting the sales of a new The starting hypothesis is that the current projections and forecasting sales, which have been done by the FIAT professional-service corporation, can successfully be corrected by the given estimation using the NAP demand forecasting model, which, ultimately, should result in more accurate forecasting.The aim of the research is to reduce uncertainty and create the preconditions for forecasting the optimization process based on the application of the NAP model through the integration and coordination f contextual information, which cannot adequately be incorporated by using the quantitative forecasting methods (primarily time series).
The application of the NAP forecasting on the example of the automobile 170 Economic Horizons (2012) 14(3), 169-179 industry could contribute to a be#ere understanding of its functioning in the global environment, especially in times of a crisis and recession and bearing in mind their interdependence.When it comes to the sales forecasting of new products, the lack of historical information favors the use of qualitative forecasting methods. The important, essential advantage of qualitative forecasting methods in relation to quantitative forecasting methods lies in their potential to forecast changes that may occur in demand for a new product, and, implicitly, in the range of its sale. Although the NAP model is based on subjective assessments characterized by a successful application in many areas of forecasting, the ability to rapidly incorporate feedback and a possibility of simple comparison to actual results .The structure of the paper is organized in the following way: in the second part which consists of two sections, a review of the literature uncovering the problem of forecasting automobile sale has been given, along with a brief description of the processed problems and the used forecast methods, as well as a review of the relevant references relating to the application of the Analytic Network Process, with a special overview on the forecasting area, whose the analyzed forecasting problem is defined.
In the third part of the paper, firstly, the NAP method is described and then applied to a specific case study. The Analytical Network Process in the literature is suggested as a solution for large, dynamic and complex problems of multiple criteria decision-making, such as he strategic planning of organizational resources, the evaluation of strategic alternatives and an opportunity of introducing new manufacturing technologies.These problems include numerous, both quantitative and qualitative factors, then many interactive a#reburies (economic, social, political, cultural, etc. ) and complex relations between them.
All these problems mostly rely on measurement and relations in the process of multi-criteria decision-making and are based on the estimation of managerial preferences. One such problem is forecasting, which includes a number of interrelated and o$en conflicted factors and appendices that need to be taken into account in order to make an optimal business decision.Finally, the model results are presented together with a postaxial analysis, as well as the conclusion, with possible indications for a future research. LITERATURE REVIEW Due to the possible implications for consumption and the economy in general, forecasting the sale of durable goods such as automobiles has a great significance, especially concerning that the automotive industry has the key role in many economies and presents their main driving force and is the generator of their economic growth and development.Demand for automobiles significantly affects the trends in travel and tourism, the development of the transportation infrastructure and residential pa#erne (Ebb Swishes & Margarine, 2002), and all these activities contribute to economic expansion and the opening of the new Jobs.
On the other hand, economic expansion puts pressure on politicians, economists, urban planners and traffic engineers, to be aware of the trends in demand for automobiles and that incorporate the feedback from them into their plans and projects. Buying an automobile is a critical consumer decision influenced factors, in both developed and developing countries.ABA Swishes & Margarine, 2002). In an a#empty to be#ere understand the movement of the automotive market and the future of the automotive industry, recent studies and the analysis cite three key factors determining the volume of automobile sale: the purchasing capacity of the population, the borrowing capacity and motivation for buying automobiles, also emphasizing the factor of the so-called restrained demand, which should be taken into account during forecasting automobile sale, especially a$ere a period of a major falling sale caused by global crises (Place, 2011).The automotive market has a large part on the racket for consumer durable goods, and companies that manufacture automobiles cannot eliminate the need to forecast the sale of new models (Kahn, 2002), regardless of the uncertainty present in their development and marketing and objective constraints, which, in addition to unrealistic expectations, o$en lead to misjudgment and great losses. In his automotive P.
Movie, Application of analytical network process in forecasting automobile sales of Fiat 500 L market research, Carlson and Mumble (1980) forecast demand for automobiles in the period from 1979 to 1983 by classifying automobiles into five categories: pub-compact, compact, intermediate, standard and luxury. The authors tried to establish the nature of the relations between gasoline prices and other relevant factors and the automobile sale, concluding that the sale of compact vehicles grew faster (from 35 to 45 percent) than the sale of other types of automobiles.They also established the fact that the economic conditions were the main determinant of a future automobile sale. The study indicated the dependence on the gasoline prices and automobile sales. However, the study was limited to two independent variables in an a#empty to forecast the sale during a difficult political period the hostage crisis in Iran and the oil embargo).
For the same reason, Harris (1986) also analyzed the impact of certain economic variables on the automobile sale and found a significant correlation between demand and some economic variables.Garcia-Ferret et al (1997), in an effort to evaluate the performance of different forecast methods. The model of the National Road Traffic Forecasts (NORTH) Roomier (1995) includes a model based on the household and explanatory models. Both models use a combination of time series involving causal variables. The multi-criteria approach in forecasting sale was suggested by Change et al (2007), wrought the development of fuzzy neural networks, Quo (2001) and Thomas & Formalism (2006), which are using classification and essence for forecasting the complexity of the environment. F these applications were dedicated to solving the problem of strategic decision-making (28%) (mainly the problem of the evaluation and selection of optimal business strategies and supply-chain strategies), and to a lesser extent to the resolving of political problems and conflicts between individual countries and companies, too.
The successful application of the Analytic Hierarchy Process (APP) and their extensions, the AnalyticalNetwork Process (NAP) in economic forecasting was demonstrated by Stats (2001), Global-Amazed (1995), Stats & Global-Amazed (1981), Blair et al (1987), Blair & stats (2010), stats (2005), Inertia & stats (2004), Y;keel (2005), (2010), vulgarity et al (2009) an so on. In support of the APP application in forecasting, the emphasized APP advantages reflected in the ease of use and a great possibility of an assessment specification, performed by the consistency test.Stats and Barras (1991) analyze the application of the APP in forecasting oil prices and forecasting exchange rates. In addition, the Analytical Network Process has proved very successful in ranking and selecting projects, as demonstrated by Made & Presley (2002), as well as Lee & Kim (2000), then in strategic decision-making, Saris (2003) and production planning Karakas et al (2002), optimal planning, Mom & GHz (2003) etc.A Review of the NAP applications The Analytical Network Process (Stats, 2001), as an extension of the Analytical Hierarchy Process (Stats, 2010), can be used in solving the problem of choosing under the conditions of uncertainty or as a forecasting the preference evaluation of the alternative courses of action, while forecasting using the APP/NAP focuses on performing the relative distribution of probable future outcomes. These forecasts are then used when the alternative courses of action are evaluated.
The review of the NAP applications published in scientific journals (Graph 1) shows that the largest number 171 Graph 1 Viewing NAP applications Source: Voluntarily et al, 2009, 40 172 good review of APP applications is given by Vida & Kumar (2006). Voluntarily et al (2009) demonstrate the use of the NAP in forecasting the sale of a new product, emphasizing the multiple criteria nature of he problem and the difference between sales forecasts in general and new product sales forecasts, which characterizes a limited amount of information, time available for an analysis and uncertainty in terms of the market response to such a new product.METHODOLOGY Analytic Network Process The Analytical Network Process (NAP) is a method for decision support, developed by Thomas Stats (2001), and allowing the involvement, quantification and objectification of all relevant, tangible and intangible factors in the decision-making process, as well as all the existing influences between decision criteria and alternatives. Archaic & Shank, 2007).
Generalizing the access of supermarkets, introduced in the APP concept, the NAP allows interactions and feedback within and between the components of the model: in clusters (the inner dependence) and between clusters (the outer dependence).This feedback successfully includes complex relations, especially in the cases of risks and uncertainties. An NAP model consists of two parts. The first part consists of a hierarchical control interactions in the studied system. The second part is the influence network , amongst the elements and clusters, whereby one NAP model can have one or ore networks. Furthermore, the problem is o$en studied through a control hierarchy or a system which consists of benefits, costs, opportunities and risk.
The synthesized results of the four control systems are combined by calculating the ratio between a product's benefits and possibilities and such a product's costs and risks in order to determine the best outcome. The procedure of applying an NAP model of decommissioning is carried out in five steps (Stats, 2001): ; the decomposition of the problem - A decision problem is decomposed into its main components. ; the cluster formation for the evaluation - A$ere fining the decision-making objectives , it is also necessary to generate clusters for the evaluation purpose by a criterion, sub-criterion (if it is possible) and cluster alternative. the structuring of the NAP model - The NAP is applied to different decision-making problems in the field of marketing, health, politics, military issues, society, predictions, etc.
Their accuracy of forecasting proved in impressive applications in the field of economic trends, sports events and other events, whose outcome became known later. ; a paired comparison and parameterization - In this step, it is necessary to compare the pairs of elements f decision-making as well as the synthesis of priorities for all the alternatives.When such a paired comparison in the NAP model is made, questions are formulated in terms of domination or an impact, which is the central concept in the application of an APP/NAP methodology. If a registry element is known, which of the two elements being compared in relation to it have a greater impact (it is more dominant) in comparison to that registry criteria? Or, in the case of an existing feedback, which of these two elements is under a higher influence of the registry criteria? The estimations are made by a fundamental scale emulate human thinking most adequately. the sensitivity analysis of the solution - It is finally possible to make a decision and carry out a sensitivity analysis in terms of the impact which, according to the importance of some criteria or subscribers, a final outcome has on a given solution ; it is also possible to determine how big or small these indicators are through an analysis.
Problem description and the construction of NAP model The model of the Analytical Network Process is applied to the problem of forecasting a sale for a new automobile model Fiat 500 L.As they say in FIAT, the del 500 L combines the inherent characteristics of the different classes of automobiles, with an aim to 173 Table 1 The scale of relative significance 1-9 Intensity of Importance Definition Explanation Equal Importance Two activities contribute equally to the objective 3 Moderate importance Experience and Judgment slightly favor one activity over another 5 Strong importance 7 Very strong or demonstrated importance An activity is favored very strongly over another; its dominance demonstrated in practice 9 Extreme importance The evidence favoring one activity over another is of the highest possible order of affirmation Mean values between two adjacent estimates When compromise is needed A reasonable assumption If activity I has one of the above nonzero numbers assigned to it when compared with activity J, then J has the reciprocal value when compared with I Reciprocals of above source: stats & Seekers, 1985, 27 offer a distinctive and versatile vehicle that would be an alternative to the traditional models of the B and C classes. The label "L" summarizes the three key dimensions, which represent a step forward in relation to the model FIAT 500: the size (Large) I. E.
A large, national and efficient space, light (Lightness), the use of friendly and ecological technologies making life easier and be#ere, and adjustable surrounding (Lo ), design allowing it to live a life to the fullest. The estimated sale of the VGA (Fiat Group Automobiles), contained in the plan presented by Sergei Maraschino, executive 2012, 3. 4 million vehicles in 2013 and 3. 8 million in 2014.
