Application of artificial intelligence using linear regression and the Naïve Bayes model in forecasting and analyzing the consumer price index in Vietnam

Abstract

The investigation of applying Artificial Intelligence (AI) and Naïve Bayes to forecast the Consumer Price Index (CPI) in Vietnam marks a significant contribution to advancing accurate inflation prediction capabilities. The study leverages rigorous methodological standards and reliable data sources by utilizing a comprehensive 2003 to 2023 dataset comprising seven input variables and the CPI as the output variable. A correlation coefficient of 0.99 indicates a robust correlation between the predicted value and the actual value. The model demonstrates efficacy in forecasting the CPI in both the training dataset and the testing dataset. Furthermore, the histogram visually represents the distribution of errors. The errors are primarily clustered at a minimal magnitude, predominantly falling within the range of -0.05 to 0.03. This suggests that the model tends to make predictions that are quite close to the actual value. The achieved Mean Squared Error (MSE) value of 0.03 demonstrates the model's remarkable accuracy, validating the effectiveness of AI in capturing intricate patterns within CPI data. This research paves the way for further exploration of advanced machine-learning techniques tailored to Vietnam's economic landscape, contributing to improved economic forecasting, proactive policy decisions, and sustainable growth.