This is an experiment to process e-commerce data. The repo contains examples of how to apply machine learning algorithms with this type of data.
The data is from a kaggle competition, you can find it here : https://www.kaggle.com/carrie1/ecommerce-data/kernels
It contians information from transactions from a UK retailer. The fields are:
- InvoiceNo: purchase identifier
- StockCode: product identifier
- Description: description of the product
- Quantity: amount of items for that product that are contained in the purchase
- InvoiceDate: date of the purchase
- UnitPrice: price of the product
- CustomerID: customer identifier
- Country: country that purchase took place
virtualenv venv
source venv/bin/activate
pip install -r requirements.txt
jupyter-notebook
The examples provided are:
- Data Cleaning: clean the data, filling missing inforation
- Data Exploration: see how the data is distributed, find outliers
- Clustering: perform clustering methods to find patterns in the data
- Classification: classify sales according to the money spent
- Recommender System: generate recommendations based on the users' history
The examples 2,3,4 and 5 use the cleaned data result from the data cleaning process.