Hybrid Smart Systems for Big Data Analysis
N. G. Kuftinova, А. В. Остроух, M. Yu. Karelina, E. N. Matyukhina, E. U. Akhmetzhanova
Abstract
In big data analysis at an enterprise, a useful option is the analysis of spatial data by machine learning, using a hybrid smart system. Machine learning permits complex nonlinear prediction with maximum precision and efficiency. A real-time big data prediction network for effective traffic flow is of great practical value. In such a network, the key problem is to construct an adaptive model on the basic of historical data. Traffic flow prediction is critical to traffic management in information technology. Real-time data proliferation led to the development of big data analysis, for which nonlinearity of traffic data is the main source of inaccurate prediction.