ABSTRACT identified topics. Key words: Artificial Neural Networks, Marketing

ABSTRACT :This aim of this paper is to give the key information regarding artificial neural networks (ANNs), With the advancement of computer technology, there has been a drastic change in management applications , to determine the effective solution is the main target in todays context. Artificial Neural Networks(ANNs) are one of these tools that have become a critical component for business intelligence. Artificial neural networks are machine learning techniques which integrate a series of features upholding their use in financial and economic applications. Backed up by flexibility in dealing with various types of data and high accuracy in making predictions, these techniques bring substantial benefits to business activities.

This paper investigates how consumer behavior can be identified using artificial neural networks, based on information obtained from traditional surveys. Results highlight that neural networks have a good discriminatory power, generally providing better results compared with traditional discriminant analysis. The purpose of this paper is to describe the basic behavior of neural networks as well as the works done in application of the same in management sciences and stimulate further research interests and efforts in the identified topics.Key words: Artificial Neural Networks, MarketingINTRODUCTION: because of Globalization competition occurs among companies and even countries.

companies need to maintain and sustain competitive advantage that can results with profitability. Marketing has an important part in competition and companies which can effectively manage costs can make more profits. Forecasting sales quantity and sales revenue is very important for a company to take action for the next period. Sales forecast enable companies to manage their budget effectively, to reduce uncertainty, to reduce risks, to speed up the decision making process etc. For this reason some techniques are used for decades. Artificial neural networks (ANN) can be defined as a highly connected array of elementary processors called neurons .

According to Specht, ANN is usually defined as a network composed of a large number of simple processors (neurons) that are massively interconnected, operate in parallel, and learn from experience (examples) . At information processing and pattern identification, ANN?s are used. It takes its roots from the working mechanism of human brains. At this model, like human brains, there are neurons which are computing units, and they are interconnected with each other in an organized fashion. Neurons process information and convert inputs into outputs. According to the relationship between these neurons, information can be generated 1.

It is a easiest form of the neural system at human brains, but if inputs have a strong connection than notable information can be achieved. One major application area of ANNs is forecasting for both researchers and practitioners. The neural network approach gives better classification, handles complex relationships better, and is stronger for interpolation. ANNs have a reliable modelling flexibility and adaptability, as they can deal the learning process. On the other hand it can adjust their parameters if new input data are available .

Using ANN in studies has benefits to some extent. Different neural system models have been created for different applications. The most important and frequent neural network is feed forward neural network. At this type there are three layers, which are; input layer, hidden layer, and output layer. At input layer, different independent variables are used to forecast output layer, which consists of dependent variable. At hidden layer, neurons are interconnected to input and output layers. This is used for examining the pattern at nonlinear relationships between output and input layers . Because at our study there is linear relationship between inputs and outputs, hidden layer is not used.

The type which doesn’t contain hidden layer in ANN is called single-layer perceptron. Inputs are simply connected with outputs via weights. According to Sexton and Dorsey, criticisms are generally focused on the inability to adequately identify identify unnecessary weights in the solution . This limits to reach a strong output but some algorithms are created to eliminate this problem. Forecasting problems occur in so many different disciplines and the literature on forecasting that use ANNs is scattered in so many diverse fields. It is hard for a researcher to be alert of all the work done to date in the area . ANNs have been used in many disciplines from the early 1980s to recent years such as engineering, medical diagnosis, data mining, and corporate business available . For example ANN is used in data mining .

There are other areas that ANN?s are successful, and these include competitive market structuring, market segmentation analysis, identification of loyal and profitable customers, also forecasting brand shares etc . ANN models are increasingly being used as a decision aid. Number of areas such as manufacturing, marketing, and retailing used it , Several authors have given comprehensive reviews of neural networks, examples of its applications, and comparisons with the statistical approach .Neural networks for marketing :Neural networks technology became a preferable destination with its influence over number of domains. Researchers designed different kinds of neural nets systems , perceptron is a feed-forward network with one layer of learnable weights connected to one or more units, which is the basic element of neural network. Perceptron is a linear classification algorithm of supervised learning. An activation function is used to reach the goal of nonlinearity3.

It combines a set of weights with the feature vector to make predictions. Perceptron can date back to the middle of last century and therefore it’s regarded as one of the earliest machine learning algorithms in the world. main areas of application of ann is in marketing research is the market segmentation.

calculated customer lifetime value, loyalty and consequently identified client segments. In most of segmentation studies auto maps were utilized and authors often stressed their advantage in interpreting the informational value of input data. They confronted self-organizing map with multilayer feedforward network and argued that map allows an intuitive representation of results, therefore is more straight forward to understand. On the otherhand, feedforward network was more powerful, since its generalization process was more robust than in case of self-organizing map. LITERATURE REVIEW : Relationship between marketing and operations, One of the first authors to discuss the relationship between marketing and operations was Shapiro (1977).

He mentioned that, in order to reduce the amount of conflict between marketing and operations, these two must understand each other’s characteristics. The marketing professionals should impart their strategies around the existing operating characteristics. It is clear that marketing people should understand operations management challenges and not only market needs.

To Karmarkar (1996), a greater interaction between marketing and operations occurs through interaction between the functions of both areas. These interactions are represented by joint decisions that can result in improved performance of factors, such as quality, lead time, cost and flexibility. For Sawhney and Piper (2002), an important interface between the functional areas of marketing and operations involves structuring and managing the production capacity. Inconsistent actions between areas of marketing and operations in terms of management capabilities result in negative impacts on delivery time, quality and cost.

According to McGaughey (1988), the market is always shifting demand towards products of ever greater complexity. This increased complexity becomes critical in order to face competition. We have observed that constant adaptation to market demands requires greater coordination between marketing and operations. According to Mollenkopf et al. (2011), operations can meet the objectives of the marketing area and offer competitive market differences, if properly designed.4 The relationship between marketing and operations has prospered in the last 20 years (Tang, 2010). Yet, there are still topics included in this interface that deserve attention. In other words, there is a need for more knowledge about the interface between the areas of marketing and operations.

One factor that requires greater knowledge about the interface between these two functional areas is the number of elements involved. The following section aims at presenting the relationship between marketing and operations, and will discuss the attributes of the marketing and competitive criteria of the operations. Relationship between marketing and operations One of the first authors to discuss the relationship between marketing and operations was Shapiro (1977). He stated that, in order to reduce the amount of conflict between marketing and operations, these two areas must understand each other’s characteristics. The marketing professionals should develop their strategies around the existing operating characteristics.

It is clear that marketing people should understand operations management challenges and not only market needs. Proposed design The design proposed in this study is based on the framework developed by Tang (2010). In this framework, the authors propose a series of decisions in the marketing area that relate to the operations area. For Tang (2010), the main objective is to attain the coordination of demand and a supply chain that maximizes profit. According to Tang (2010), from the moment that the marketing and operations areas conduct their activities together, the company obtains better performance .

The proposed design is based on the areas not highlighted in gray.4 It should be noted that Tang’s approach (2010) is different from that of Shapiro (1977), which defines the existence of possible conflicts between marketing and operations due to the characteristics of their responsibilities. For Shapiro (1977), it is not possible to eliminate conflicts, but they can be minimized through greater knowledge, Our model attempts to unify these two approaches: (i) a line of study proposed by Tang (2010) that states that the decisions of the marketing field can be synchronized with the operations area in order to seek better performance, (ii) and a line of study proposed by Shapiro (1977) which states that there is no way of avoiding the conflict between marketing and operations, requiring assimilation of knowledge about this relationship in order to keep track of activities and to minimize conflicts. The input variables are grouped according to the framework proposed by Tang (2010): product, service, quality, price, promotion, and delivery channel. The output variables are grouped according to coordination and collaborative demand, since the objective is to achieve the best delivery performance, that is, on-time. Tang (2010) ranked different marketing decisions and operations performance in these categories.

The relationships among variables presented in our model have the intention of representing those in Shapiro’s study (1977). According to this author, there is an interface between marketing and operations activities that relate to one another. There are a number of relations among these decisions, also called processing activities, that impact the outputs of our model, which constitute the delivery performance of the operations area.

In short, marketing decisions are the input variables of the model, and the processing stage of these decisions determines the relations between them, while delivery performance is the output variable. conclusion Results indicate that the ANN enables efficient management of investments. The preparation and establishment of the model of a distinguished group of 6 high-risk factors: financial, scientific-technical, manufacturing, company, market and external-the eco. Within each group, the identification of risk factors 5.

There were three clusters are : aggressive, conservative and moderate. The aggressive risk cluster of companies 30 different risk factors to the ANN model is able to include all the factors. The conservative risk cluster of companies to the ANN model could add only 12 risk factors. The remaining 18 risk factors must be eliminated. The moderate risk cluster of companies may include 19 risk factors. 11 risk factors must be eliminated.

Thus, every company must know what level of risk the company may assume by investing in technological innovation using ANN modelReferences 1. Rytis Krušinskasa , Management Problems of Investment in Technological Innovation, Using Artificial Neural Network, 20th International Scientific Conference Economics and Management – 2015 (ICEM-2015) 2. Michal , Robert Verner ,Artificial neural networks in business: Two decades of research, Applied Soft Computing3.

Ashkan Zakaryazad, Ekrem Duman n, A profit-driven Artificial Neural Network (ANN) with applications to fraud detection and direct marketing, Neurocomputing4. https://pdfs.semanticscholar.org