Sales forecasts help make informed decisions about everything from staffing and inventory to new product lines and potential marketing efforts. Sales forecasting allows sales managers and reps to spot potential issues and gives you time to avoid or alleviate them.

Here, Machine Learning project was developed for predicting sales Delivery Club. The predictions were made with classic ML models (Random Forest, GradBoost, Linear Regression, etc.).

Dataframe contained information about the date, weather conditions, product types, stores, and their locations. The history of the sales for previous 7 months is available for delivery shops in 10 cities.

• sales_forecasting.ipynb - Main notebook with description of the project, code, graphs, and comments.
• sales_forecasting_supplementary.ipynb - Additional code needed to convert the notebook to web-page.
• data - directory, that contains dataframes, used in the project.
  • train.csv - train dataframe, used for training ML models.
  • test.csv - test dataframe, used for predictions.
  • public_data.zip - contains two dataframes (train.csv and test.csv) in a single zipfile.