CO-MOVEMENT CLUSTERING: A NOVEL APPROACH FOR PREDICTING INFLATION IN THE FOOD AND BEVERAGE INDUSTRY

Abstract

In the realm of food and beverage businesses, inflation poses a significant hurdle as it affects pricing, profitability, and consumer’s purchasing power, setting it apart from other industries. This study proposes a novel approach; co-movement clustering, to predict which items will be inflated together according to historical time-series data. Experiments were conducted to evaluate the proposed approach based on real-world data obtained from the UK Office for National Statistics. The predicted results of the proposed approach were compared against four classical methods (correlation, Euclidean distance, Cosine Similarity, and DTW). According to our experimental results, the accuracy of the proposed approach outperforms the above-mentioned classical methods. Moreover, the accuracy of the proposed approach is higher when an additional filter is applied. Our approach aids hospitality operators in accurately predicting food and beverage inflation, enabling the development of effective strategies to navigate the current challenging business environment in hospitality management. The lack of previous work has explored how time series clustering can be applied to support inflation prediction. This study opens a new research paradigm to the related field and this study can serve as a useful reference for future research in this emerging area. In addition, this study work contributes to the data analytics research stream in hospitality management literature.