Cognitive Graphs Preventive Analysis on the Example of the Consumer Price Index Forecasting Model

The economic crisis caused by the pandemic has demonstrated the inconsistency of many predictive models. The relevance of the work is related to the high frequency of shocks that occur in the Russian economy, as well as the lack of preventive methods that can improve forecast models. The purpose of this study is to test the correctness of the well-known econometric model in an economically unstable situation and to strengthen it with the help of a cognitive graph model. The task of predicting inflation without exaggeration is very important for government institutions, businesses, and households. The accuracy of the forecast obtained directly affects the efficiency and consequences of decision making in monetary policy, in planning expenses and savings of households and in business financing. Research objectives. Build and describe a model of interfactor relationships of the inflationary model, check the model for balance, demonstrate the failure of the forecast in conditions of instability, describe a method that strengthens the original model. The cognitive graph method is used as a warning tool to improve the econometric inflation forecasting model. As a confirmation of the inefficiency of the original model, the results of inflation assessment for the COVID time interval are provided. Conclusion. Economic systems modelling using cognitive graphs is an effective preventive method that allows you to dynamically monitor the internal cycles of the system and points out key factors which dynamics should be closely monitored. In addition, the method of cognitive modelling can determine the key factors of the system, exclusion of which at some point will necessarily lead to incorrect results of the model. An algorithm for constructing a cognitive graph on the initial econometric inflation forecasting model is demonstrated.

inflation, cognitive graph, consumer price index, CPI, regression, forecasting