Early warning prediction system for high inflation: an elitist neuro-genetic network model for the Indian economy

In this paper, we propose elitist genetic algorithms–based artificial neural network (ANN) model for setting up an early warning system for occurrence of high inflation. The proposed warning system uses values of an appropriate set of economic fundamental variables as input and builds an ANN model for quantifying the possibility of high inflation within a fixed period of time window. Elitism-based generational genetic algorithm is used for optimizing the architecture of the ANN model. We empirically evaluate the proposed neuro-genetic approach to identify the class of leading economic indicators and build an early warning signalling system of an occurrence of high inflation (overall and component inflations) using the data from the Indian economy. We further compare the results of the proposed approach with the commonly used data-driven signals approach. In the empirical studies, we observe promising performance of the proposed neuro-genetic warning system, which is capable of generating accurate early warning signals of an impending high inflation.