Grouping cities based of their socio-economic indicators

In this paper a new approach in analyzing city behaviour and its dynamic developments is proposed. Usual attributes are divided into two larger groups: social and economic. This idea makes possible analyzing relative performances of cities from different angles, such as: two-dimensional clustering; city ranking by simple calculating their distance from origin; minimum sum of squares clustering, etc. Such an analysis is helpful to show good and week points in developing city strategy in long term period. Grouping of 77 capital cities of regions in Russia is considered. The averages of both groups, for each city are calculated, covering 12 years. Minimum sum-of-squares criterion is used for clustering and solved by variable neighborhood search. It appeared that there are a few outlier cities, two of them being non dominated or efficient. Detailed comparative analysis of the results before and after crisis (economic sanctions) are also provided.