Page 22 - Azerbaijan State University of Economics
P. 22

Safwan Al Salaimeh, Eman Al Alreyati:  Using of New Information Technology in the Creating of
                                                                     Promotion Products

                                                      th
                           (  ) − actual absolute values i characteristics of alternative x;
                          

                                                            th
                        Kinx, kinl – the best or the worst value i  characteristics on a variety of alternatives
                       on which the choice is made.

                        ∝ -  Nonlinearity parameter, determining the nonlinearity of the dependence of
                           
                       utility value on criterion value [Safwan Al Salaimeh, Khaled Batiha, 2006; Safwan
                       Al Salaimeh, Amer Abu Zaher, 2011. Safwan Al Salaimeh, Pushkarev A.N.,  2011]

                       After the structural identification on the parameters of the models are determined;
                       coefficient  value  ai  and  nonlinearity  parameter  αi.  at  the  same  time  objective
                       approaches to assess the individuals references based on the result of fixed act of
                                                           th
                       choosing an alternative choosing an S  alternative means, that its utility is superior
                       to the utility of any other alternative

                                                   (   ) >   (   ),         = 1,   ;
                                                               
                                                       

                                                        (   ) >   (   );
                                                                     
                                                            

                                      
                                              
                          (   ) =       +       + ⋯ +           >           +           + ⋯ +           =   (   );
                             
                                                                                                    
                                                                                             
                                                                         2   2
                                                                1   1
                                                               
                                         2    2
                                          
                                       
                                                                                       
                                                                                
                                                
                                                                 
                                    (     1  −       ) +    (       −    ) + ⋯ +    (         −         ) < 0
                                                                            
                                                     2
                                                          2
                                           2   2
                                  1
                                                                 2
                                {                             ⋮
                                                              − 1

                                                               1
                                                          
                                               (       −    ) =    ;                 ;
                                                          1
                                                   1
                                                                 2

                       Then the identification model can be represented in the following form;

                                                                     ∑           < 0;           = 1,   ;
                                                      =1
                                                             

                                                                      ∑           = 0,      = 1,   ;
                                                      =1
                                                                

                                                                    0 ≤    ≤ 1;   ∑        = 1;   ∀   = 1,   ;
                                                                        
                                                         
                                                                   =1

                    To determine the point estimate i.e specific numerical values of the coefficient ai, the
                    approach based on definition of a chypyshal point, it is called chypychal approximate point
                    and corresponds to the solution which in the great degree satisfies the given organ:

                       =                ∈Ω    ;  where L – distance from any point x to constraint line; Ω – the
                    range of possible values  of the coefficient ai [Safwan Al Salaimeh, Zafer Makadmeh,
                    Avramenko V.P. Shtangee S.V., 2012; Safwan Al Salaimeh, Mohammad Bani  Younes,
                    2014; Safwan al Salaimeh, Zeyad Al Saraireh, Jawad Hammad Al Rawashdeh, 2015]

                                                           22
   17   18   19   20   21   22   23   24   25   26   27