The dynamics of psycho-physiological characteristics of participants of laboratory markets in the process of making economical decisions in the decentralized control system is analyzed. For this purpose, the method based on the comparison of stabilographic data with the history of market actions recorded in the course of the experiment is used. The key tool for data analysis is the new segmentation algorithm, which provides efficient partitioning of the stabilographic time series into homogeneous fragments. The segmentation algorithm is obtained as the solution of the problem of estimation of the parameters of the hidden Markov model. The application of this algorithm on the level of individual decision making proves the hypothesis of connection of the stabilogram segmentation time instants of the participant with the time instants of signal actions on the laboratory market. On the level of group decisions, the effect of synchronization of stabilographic time series of participants at the time instant of auction culmination connected with revelation of private information is supported. The degree of synchronization is estimated using a proximity factor calculated based on the specially aggregated canonical correlation.
|Number of pages||18|
|Journal||Journal of Computer and Systems Sciences International|
|Publication status||Published - Dec 2009|