The article is devoted to solving the problems of data mining on the investment development of regional socio-economic systems. A procedure for data mining is proposed, which is based on the application of the principal component method and the construction of self-organizing Kohonen maps. The decomposition of features characterizing the production, investment and social aspects of the state of regional socio-economic systems has been performed. An algorithm for the intelligent analysis of samples of the lower and middle decomposition levels has been developed. The algorithm assumes multiple component analysis for samples of different decomposition levels. The component and neural network analysis of data on investment processes in the regions of the Russian Federation has been performed. The results of the analysis are presented in the form of cartograms of various levels of generalization, on which clusters of regions are identified and their distinctive characteristics are formulated. The obtained data mining results provide support in applying a creative approach to solving the tasks of managing the investment development of regional socio-economic systems.