Store Data Time Series Analysis

This time series analysis is conducted on sales data retrieved from kaggle. Initial analysis attempts to understand what type of product is the most popular. Analysis shows that products labeled as GROCERY I is the most bought out product at the store followed by beverages and produce. Different bar charts and pie charts were also used to understand the monthly average sales and sales per store types - Sales are higher in December and type A and D stores have the highest amount of sales.

A time series analysis were performed with train data to predict GROCERY I in the later part of my analysis. Preliminary steps including stationary test and seasonality removal is taken prior to modelling. ARIMA(5,0,5) was selected due on its higher AIC score. Forecast suggests that sales for GROCERY I will slowly decrease and fluctuate less and less at the beginning of 2017.

Click to see the analysis and results