Machine Learning Based Restaurant Sales Forecasting

To encourage proper employee scheduling for managing crew load, restaurants need accurate sales forecasting. This paper proposes a case study on many machine learning (ML) models using real-world data from mid-sized restaurant. Trendy recurrent neural network (RNN) are included direct comparison to methods. test the effects of trend and seasonality, we generate three different datasets train our with compare results. aid in forecasting, engineer features demonstrate good methods select an optimal sub-set highly correlated features. We based their performance forecasting time steps one-day one-week over curated dataset. The best results seen come linear sMAPE only 19.6%. Two RNN models, LSTM TFT, ensemble also performed well errors less than 20%. When one-week, non-RNN poorly, giving worse 20% error. extended better scores 19.5% result. overall seasonality removed, however simpler ML when linearly separating each training instance.