#R #TimeSeriesForecasting #Prophet #RShiny
Project Report - https://www.dropbox.com/s/b3eutee5lojlfag/Final%20Project%20Team%209.docx?dl=0
Project Video - https://drive.google.com/open?id=1Gv-APge2tJCMaJXZwuYh07jcTHRDx8Uo
We built various demand forecasting models to predict product demand for grocery items using R Packages. It is a timeseries analysis problem and there are various methodologies to solve these kinds of problems. Time series forecasting is a technique for the prediction of events through a sequence of time and it is technique to predict future events by analyzing the trends of the past, on the assumption that future trends will hold similar to historical trends. The methods we’ve employed our model are ARIMA, XGBoost , ETS (Exponential Smoothing State Space Model) and Prophet .
Comparisons were made along the following dimensions:
- Predictive performance
- Scalability
- Ease of use scalability
We have used R Studio to code the model. On the basing weightage on the above mentioned comparisons we chose the best time series modelling techniques and predicted the demand values for the next two weeks.
The data associated with the project can be found at the following link: https://www.kaggle.com/c/favorita-grocery-sales-forecasting/data