Complex-valued economics in predicting one dimensional economic series

Abstract

Purpose: аn article prepared based on the results of the RFBR grant No. 19-010-00610 \ 19 «Theory, methods and techniques for predicting economic development by autoregressive models of complex variables» is devoted to the formation of a new set of methods of economic forecasting using a complex-valued economics. Discussion: еconomic processes are diverse, which has led to the emergence of many methods and models for forecasting these processes. These methods are being developed and improved. One of the new approaches to improving forecasting accuracy is the use of complex-valued economics methods to solve this problem. The article discusses possible options for using complex-valued models for short and medium term forecasting. Results: the author proposes new models for short- and medium-term forecasting, which have new properties than the model of real variables. It is proved that in some cases these models are more accurate in economic forecasting than the existing ones.