This paper aims to construct a model and algorithms that apply econometric and neural analysis to identify major factors influencing carbon dioxide emissions. We use the database Enerdata which collects global data on the energy industry, from which data on the nuclear power of different countries are extracted. The computational results obtained in the linear analysis and neural analysis of such parameters as crude oil production showed a degree of influence of 1.16% (in the linear analysis) and 24.13% (in the neural analysis).