The trend analysis is designed to study changes or “drift” of the local average value of the time series with the use of a mathematical model and further forecasting on its basis. Accurate and timely analysis of future events helps make clear and proper management decisions that will allow companies and regional actors to take their rightful place in a market economy and keep their competitiveness level. The trend gives an opportunity to make short or long term predictions in the shortest possible time with little available information. The article discloses the application of 3 simplest econometric trend models: linear, quadratic (polynomial) and exponential on the example of innovative activity of RF organizations. It presents the process of analysis and evaluation of projected trends with the use of statistical tests indicating the adequacy and reliability of the constructed models. On their basis the paper makes a forecast for the near future, and identifies those mathematical tools ands that can be applied in further research. For a more accurate assessment of innovation activity the article presents the similar trend models for 2 RF subjects: the Vologda Oblast and the Perm Oblast. It shows mechanism of selecting the best trend of the constructed ones. It makes a conclusion about the current state of innovation activity in the country, lists a number of problems that have a negative impact on the innovative activity level and ways of dealing with them. The main methods include tax incentives and government subsidies. The author indicates that the significance of innovation should be determined not only by how they meet new people’s needs, but to what extent they serve the old ones. The need to conduct innovative activities should be built in at the early stages of the educational process, as innovation is one of the pillars, the stable development of economy rests on
Keywords
dynamics, forecasting, innovative activity, trend, time series