Optimization of Restaurant Operations and Food Waste Management Through Day-Specific Sales Forecasting Using ANFIS

The inability to accurately predict daily sales hinders restaurant managers from efficiently managing food ingredients and raw materials. Accurate sales forecasting enables better control over stock levels, reducing food waste and ensuring the timely use of perishable products. This study applies an Adaptive Neuro-Fuzzy Inference System (ANFIS) to predict day-specific sales patterns for the restaurant TB2, taking into account factors such as holidays, seasonal trends, and marketing campaigns. Main effects plots generated using Minitab showed that Saturday had the highest dine-in sales, while Sunday recorded the highest online and total sales. Furthermore, festive seasons were found to enhance both dine-in and online sales volumes, while advertisements and public holidays had a significant positive effect on dine-in sales but only a limited impact on online sales. The accuracy of the ANFIS model was evaluated using the root mean square error and the coefficient of determination. The root mean square error between the predicted and actual total sales was 4589.21, with a coefficient of determination of 0.804, indicating strong predictive performance. The ANFIS model proves effective in forecasting restaurant sales for any given future date, enabling better operational management and food waste reduction.