Objective: The objective of the project is to analyse the revenue for kids store and forecast for the next two years on quarterly basis.
Project Approach in Brief:
- Created baseline naïve models to understand underlying patters in the data.
- Built ARIMA and Holt Winter triple exponential models to forecast revenue.
- Performed statistical comparison (RMSE, MAPE) for all the techniques and provided business recommendations.
Project Sneak Peak:
Moving averages gives overall trend based on historical values in a given time frame.
Naive forecasting techniques (naive, seasonal naive, mean) provide benchmarking results againist sophisticated models such as ARIMA, SARIMA, ETS and Holt Winter.
For full project report check out the Kids Revenue Forecasting Final Presentation
Tools Used: R Programming
Libraries Used in R:
- forecast
- TTR
- tseries
- ggplot2
Licensed under the Apache License 2.0