Today the trend in regional politics and many state programs for regional development in Ukraine is largely based on approaches that have been used for leveling disparities in socio-economic development of the former USSR. The primary place they occupied the processes centralized fnancing of the economy. To some extent, these measures can smooth inter-regional socio-economic differentiation, but they can not eliminate the causes of the problem, and as a result it leads to passivity particularly underdeveloped and depressed regions. This multifaceted study will understand regional processes of Dnepropetrovsk region in general and its particular areas. The purpose of this study is to improve the analysis of behavior in the dynamics of regional economic means of mathematical modeling. The paper proposed to apply self-organizing maps of Kohonen tools that will not only determine the location area among others, but also to understand trends in its behavior. In contrast, cluster analysis, which allows only group homogeneous groups of indicators procedure that is performed by using self-organizing maps of Kohonen, except that groups the indicators also helps to visualize the results, thus transferring the problem of multidimensional space in two-dimensional space of dimension. Testing of the proposed approach was performed on the data area of the Dnipropetrovsk region for eight years (2005- 2012 years.). The information base was formed on the basic data on the site of the Main Statistical Offce in Dnipropetrovsk region. Analysis of the results confrmed the fragmentation of regional space Dnipropetrovsk region. Promoting prudent public policy at the regional level will help align the situation in the region and will clearly understand the dynamics and trends. The calculations are made to the fnancial and economic indicators Dnipropetrovsk region confrmed the adequacy of the approach. Further research should be conducted towards the incorporation of quality indicators, the main of which is the standard of living.