Research of time series

Research of time series

The study of temporary (dynamic) ranks allows us to consider and explain the development of the economic process in time, to identify positive and negative trends that are characteristic of it, to pay attention to the most important of them, and often, predict the development of the process in the future. The analysis of the dynamic series is usually associated with the forecasting of certain indicators, since the existing objective trends of changes in economic indicators to a certain extent determine their value in the future. This explains the close attention paid to this method. To identify trends, fluctuations and other basic characteristics of the dynamic series, in order to analyze on this basis, some methods of mathematical and statistical processing of the rows and primarily methods of equalizing dynamic series are used to eliminate random deviations on this basis.

During a comprehensive analysis of the development of construction and its material and technical base in the region, this method is one of the most important, used in combination with the balance method, groups, comparisons, average, etc. If necessary, a detailed study of the development of any process over time, analysis, of course, must not be carried out for five-year, but for a longer, for example a ten-year period. Thus, to confirm the uniformity of the source data, consideration of the actual cost of construction and installation works performed by large territorial leaders or construction departments for ten years is consideration, which allows us to identify those crucial years in which the most significant changes have occurred in the cost of costs of costs, and connect these changes with the factors that caused them (for example, the excess of the actual cost of the main wage in the structure of the cost of work compared to the planned can be caused by a change in the demographic situation in the region). In addition, it is often necessary to consider several interconnected dynamic series, in which level changes do not necessarily coincide in time (one phenomenon can occur with some lag in relation to another phenomenon associated with it).

In the study of such time series, when it is required to determine how much fluctuations in levels of one row depend on fluctuations in levels of the other, use the methods of correlation and regression analysis. However, before applying these methods, it is necessary to establish a logical connection between the indicators under consideration!