Restaurant Visitor Forecasting
Predict restaurant visitors using linear regression by R & MySQL. In this report, first the required data set was extracted from a SQL database using join commands. Then started with exploratory analysis to identify, evaluate and understand the main variables. Furthermore, a linear regression method is carried on to forecast the visitors based on the provided time series. At last the prediction was tested with multiple approaches to develop the most efficient and accurate strategies for restaurants to initiate planning for proper resources and services. Final result: R^2 = 0.55 , RMSE = 12.27
More Feature Explorations: Number of visitors to a restaurant certainly depends more things on its reservations, genre or area. There could also be a correlation between the social-economic conditions of the neighbourhood, other attributes like the vibe of the restaurant itself and the number of visitors.
Times Series Models: By now the visitor's data is available on a daily basis, but if it will be available on an hourly basis then we can further reduce the scale of prediction to next hour visitors prediction. It will give more control to restaurants in terms of being prepared for the next hour rush and thus further improvement of productivity and food delivery time.