Time Series Forecast of Sales
- Data was sourced from Kaggle and contained over 900,000 instances
- Explored the dataset and discovered data were grouped by store type and item type
- Gave insights on items that are selling bad and contributing less to the sales return
- Gave insights on the seasonality of the trend which helps business owner make informed decision
- Built a reusable function that can be tested on the store type, item type and range of year required for analysis
- Models used include ARIMA
- Skills used are Python, Matplotlib, Sklearn, Numpy and Seaborn.