DECISION SUPPORT SYSTEM AS A TOOL FOR THE FORMATION OF EFFECTIVE COMMERCIAL ACTIVITY ON THE BASE DATA MINING

DECISION SUPPORT SYSTEM AS A TOOL FOR THE FORMATION OF EFFECTIVE COMMERCIAL ACTIVITY ON THE BASE DATA MINING

Yernazar Rakhimzhanov

Student, Almaty Universityof Power Engineering and Telecommunications,

Kazakhstan, Almaty

Feruza Malikova Feruza

Professor PhD, Almaty Technical University,

Kazakhstan, Almaty

ABSTRACT

This article describes the algorithm of the decision support system. Factors that may influence the choice of methods for intelligent decision support system are described. A method called Data Mining (DM) was chosen for the study. The process of generating possible solutions is also investigated. The system was seen as an effective tool in business for the manager to make subjective decisions using objective data provided by the system. The scheme of actions of the process is illustrated, when the decision is made not individually, but through negotiations, which are called the negotiation support system (ATP).

Keywords: DSS, DМ, generation, algorithm, matrix, IDEF0 diagram, interfa, grammar, (NSS).

Decision making is a daily human activity. Obviously, in the decision-making process, performing the actions we are accustomed to, we do not even think about. If we talk about it more, the situation becomes more complicated. In our rapidly changing world, it is difficult to work with a lot of information. When making priority decisions, it is very important to analyze the data when, for example, it is a company that strives for stable growth. In this aspect, computer technology is increasingly used to solve management problems. Most commercial, public and government organizations no longer make serious decisions without the use of computer analysis. To this end, many decision support systems have been developed that can generate possible alternative solutions. In this paper, the decision support system is described by the method of "data mining", which was chosen as an effective method that allows you to make decisions using several approaches.

Decision support system (DSS) is a method that helps to issue the correct option and in this method, of course, there is the presence of artificial intelligence (AI). [1].

1. Definition of the problem (situation);

2. Determining the scale of this problem;

3. Formation of possible actions to solve the identified problem, depending on the level;

4. Evaluation of the formed algorithm of actions;

5. Amendment if necessary;

6. The decision process;

7. Evaluation of the work done.

This algorithm contains a list of correct actions that help solve the problem. Here, in addition to identifying the problem, the identified problem is evaluated on a scale and its characteristics to make appropriate effective actions and decisions.

There are factors that can influence the choice of methods on which a computer decision support system will be based. In Figure 1 we can see some of them.

Figure 1. Factors influencing the choice of computer decision support methods[2]

To solve the problem it is very important to find out the created algorithms. As he says, “the first wash, and what ironing. But not a refusal. " The support algorithm, which is part of the steps generated by the IDEF0 diagrams, is usually supported. IDEF0 is a convenient way of functional and graphical modeling. With this scheme, you can write a business process, referring to the logical connections between the works. A distinctive feature of IDEF0 is its emphasis on the subordination of objects. Figure 2 shows an example of such a graph.

Figure 2. Decision-making algorithm [3]

This is an algorithmic production that takes place in the process of execution in the archaic sequence. In this case, the process of repetition is possible, but still not by the mystery of a certain path. If you fall, this process is similar to life: when a person does not pass the exam for his life, he is not in the second stage of his life. One of the essential distinctive abilities of a person stops and opens his business, adding an exam and moving on to the next stage on the algorithm of life.

Based on this, it is possible to represent the structure, composition and interaction of individual DSS blocks in the form of a scheme, which is illustrated in Figure 3.

Figure 3. The structure of a computer decision support system [4]

To make a decision, you first need to identify the problem, analyze the scale of the problem, assess the situation. To cope with this, many techniques have been developed for analyzing the current situation, such as: data Mining, Data Mining, On-Line Processing - OLAP (operational data analysis), Knowledge Discovery or Intelligent Data Analysis.

Data mining reveals hidden trends, patterns, correlations between objects or events when processing algorithms. To make a qualitative decision, it is necessary to take into account all possible influences, which is what the IAD does. All methods are programmed in the IAD, which are a logical generalization of various analytical approaches.

For example, consider the decision-making process in business. In business, it is very important to research and analyze the level of sales of goods. At the same time determine which type of product is well bought and which is not sold. Identify the reasons if the product is not sold for more than a month. For example, in a supermarket where there is a large volume of goods, their location is very important. The product cannot be sold simply because it is invisible. Or simply irrelevant. To understand this, we need to gather the full picture. That is, a picture is collected from a study of the sale of individual goods. The task of DSS in this case is to restore the full picture of the state of sales on the basis of these fragmentary data. An example of this is Figure 4 below. Looking at this cube, you can understand in which region, which product fell and organize the following algorithm.

Figure 4.Type of information cube «Sales volume» [5]

In the work of Trachtenhertz, it is said that the data obtained after investigating the situation is added to the system. For example, Figure 4 highlights an object identified as a suspected object with a reduced sale. This object is identified using the formula given below, which allows you to calculate the expected value of the parameter under study using the coordinates of the point. And also, after identification, determines the accuracy of this assessment [3].

                                                      (1)

Where -relative coordinates of a point inside the corresponding cell,  – field values at the point - coefficients depending on the values of the field in the grid nodes.

As mentioned above, in order to start taking action in order to solve problems in the field of reducing sales of a certain type of product, it is necessary to determine how much has been reduced, what other threats of reduction may be, as well as the scale of this problem. To determine this , the system must solve the problem using the following formula:

                                             (2)

 - the value of the field in the jth node of the grid, -the measurement value at the i-th point, - the "weight" of the value at the i-th point, -distance from the i-th point to the j-th node, d - grid pitch, k –dimensionless parameter set by an expert. Averaging is performed based on the results of measurements at points lying inside the rectangular grid cells closest to the node to a depth of k.

Data analysis using a supporting system is based on three approaches:

1.   An expert system in which the experience of the manager is recorded to assess the situation.

2.   An analogical approach in which the analysis process takes place based on the experience of decisions made in the past. That is, the situation is evaluated by analogy with the assessment of the previous situation, even if they have no relation to each other. It's like evaluating the new according to the criteria of the old.

3.  Combinational, where the results obtained by analyzing historical data are evaluated based on the expert's experience. This is a great opportunity for the manager to show his expertise in order to evaluate the new according to the criteria of the new.[6]

As in data analysis, the generation of solutions can be carried out in two ways:

1.   To make unexpected, fundamentally new decisions that the computer is not yet capable of;

2.   Solutions based on typical scenarios, by analogy, based on a combination of known partial solutions; generation of such solutions is available to a computer.

     (3)

    (4)

   (5) 

             (6)

               (7)

              (8)

After the dialogue, a scheme of possible countermeasures is built, where the main task is to choose the best solution in the current situation. This scheme can be seen in Figure 6. According to the scheme, you can see the following algorithm of actions: after detecting a decrease in sales, the command center is activated, that is, taking measures against the identified problem. From the bottom and on the right side of the arrow there is an assessment of the situation at the expert meeting. B and F intersect in the arrow that leads to the purchase of new funds, which can be identified after D. At the same time, H has a node in the arrow that leads to C, which determines the action after the purchase of new funds - the reorganization of the composition. In the scheme I, J, L, K have a correlation, in addition, J and K are, as it were, feedback for I and L. Well, it is clear that for the implementation of I, the implementation of G is paramount. If process K gives a good result, then the execution of process M and E will be as silent [7-8].

Summarizing the process of this scheme, we can say that the system is structured so cleverly that when performing the proposed actions, the result will be a side effect of the correct organization.

Figure 5.The scheme of possible countermeasures, built by the DSS based on the results of the dialogue, to eliminate problems in the flower shop

By describing the process using such a grammar, you can get all the valid scenarios, from which the DSS will subsequently choose the best one [9].

Since the negotiation process is entered on the basis of individual decisions, there is a phase in the negotiation support system between an individual and a quantitative decision. The negotiation process takes place after the end of the first phase. For clarity, Figure 6 shows the structure of such a process.

Figure 6. Alternating phases of individual decision-making and negotiations [10]

If you need to negotiate and make deals, take action. Such a need may arise as a result of an attempt, such as a fall in production, but may be constant, for example, when buying and selling goods.

In conclusion, in our time, digitalization is not surprising, given the importance of human activity, you can get to artificial intelligence. These elements of intelligence help humanity to receive more information, and what is not collected, necessary, is not lost in such information when it is necessary to analyze data, statistics and other tasks. Given that a maintenance system is supported, including possible identified and their problems, this system should remain a tool chosen at a distance. When we see a problem for something impossible, we see only a fragment. It only takes a little time to see the whole picture, and it will happen where you need to influence. The system is only a tool that estimates our distance in technology to each case, and with the help of objective systems - when changing a new subjective system. The human factor is inappropriate, depending on how the system analyzes the increase in data, and can be realized using this subjective method of generation and valuable choice, using your design style and using your design style now. -production.

References:

1. Trakhtengerts E.A. Computer systems for supporting managerial decision-making, Management problems, 2003, issue 1, 13-28.
2. Trakhtenherts E.A. Possibilities and realization of computer decision support systems // News of the Russian Academy of Sciences. Theory and methods of management. - 2001. - No. 3. - pp. 86-113.
3. Trakhtenherts E.A. Computer support of negotiations in the coordination of management decisions. - M.: SINTEG, 2003.
4. Belenky A.G. Fedoseeva I.N. Forecasting the state of dynamic complex systems under uncertainty. -M.: VC RAS. 1999
5. Trakhtenherts E.A. Computer decision support. -M.: SINTEG, 1998.
6. Eom S.B. Decision support systems research: reference disciplines and a cumulative tradition // The international Journal of Management Science, -1995. -Vol. 23. No. 5. -P.511-523.
7. Methods and models of data analysis: operational and mining / A. A. Barseghyan, M. S. Kupriyanov, V. V. Stepanenko [and others]. SPb: Peterburg, 2004.
8. Petersen, A.H., Ekstrom, C.T. Your assistant for documenting supervised data quality screening in R/ A.H.Petersen, C.T.Ekstorm // Journal of Statistical Software. – 2019. – p.90.
9. The intelligence of the modern enterprise / S. D. Nikolaev, A. V. Zaitsev, V. V. Baranov, Th. Kraft. M. Publishing house "Komsomolskaya Pravda", 2010.
10. Baranov V. V., Baranov I. B. CRM-system as an element of the logistics chain of entrepreneurial activity in the field of high technologies // Logistics. 2016. № 10.