:  Novikov D A
:  Incentives in organizations: theory and practice
:  
( ):  Proceedings of 14-th International Conference on S
:  2001
:  
:  Proceedings of 14-th International Conference on S
:  Novikov D.A. Incentives in organizations: theory and practice / Proceedings of 14-th International Conference on Systems Science. Wroclaw, 2001. Vol. 2. P. 19 29.
:  The paper contains the consideration of the modern state of art of incentive models and their applications. In the framework of the management theory incentive problems are mostly studied for simplest organizations which consist of one principal and one agent. Game-theoretical problem is for-mulated as following. Agent's goal function reflects his preferences over the set of feasible actions and depends on his own strategy and on the strategy of the principal. The game solution set is defined as the set of the equilibriums under the given management. The management problem is to maximize the guaranteed efficiency of man-agement. As the basic model, briefly described above, considers the simplest organization intellectual active system (IAS), which consists of one principal and one agent, who make their decisions under complete information, the extensions of the basic model are generated by complicating sequentially the model. Any IAS is described by the following parameters: elements of the IAS (principals, agents, etc.); its structure (the set of relations between the elements); feasible sets of elements strategies; goal functions of the elements; information possessed by the elements at the moments of their decisions making; the sequence of getting the information and making decisions. Thus main extensions of the basic model are: multi-agent IAS; dynamic IAS; multilevel IAS; IAS with distributed control; IAS, which operate under uncertain environment or/and incomplete information. The main problem, which arises when applying results of theoretical investigation in practice, is the problem of IAS identification. Nowadays several levels of detailing exist. The most general one is to identify classes of possible values of IAS parameters. Sometimes such a general information (for example, corresponding to the assumptions of convexity of sets and functions in the model) is sufficient for restricting the class of optimal control variables. The next stage of detailing is the introduction of certain assumptions about classification of the models parameters. And, at least, the last stage is to assign certain numeric values to the parameters of the model. This stage, being based correctly through natural or simulation experiments, allows to make reasonable conclusions about the optimal control.
:   ( )
:  
:  
:  

: (pdf)

: 5609, : 912, : 6.


© 2007.