E-commerce User Analytics & Purchase Prediction

This project analyzes customer behavior using the Online Retail dataset from the UCI Machine Learning Repository. The dataset covers transactions from 01/12/2010 to 09/12/2011 for a UK-based online retailer.

Project Components

1. Data Preprocessing:

   • Cleans and processes the raw data.
   • File: scripts/preprocessing.py

2. Exploratory Data Analysis (EDA):

   • Analyzes sales trends, top products, and customer behavior.
   • Notebook: notebooks/EDA.ipynb

3. Predictive Modeling:

   • Builds a Random Forest model to predict customer purchase behavior.
   • File: scripts/model_training.py

4. Dashboard Visualization:

   • Interactive dashboard built with Streamlit to display analytics.
   • File: scripts/dashboard.py

5. API for Recommendations:

   • Simple API using FastAPI that provides product recommendations.
   • File: scripts/api.py

How to Run

1. Clone the Repository:

   git clone https://github.com/YOUR_USERNAME/ecommerce-analytics.git
   cd ecommerce-analytics
   

2. Install Dependencies:

   pip install -r requirements.txt
   

3. Data Preprocessing & Model Training:

   python scripts/model_training.py
   

4. Run the Dashboard:

   streamlit run scripts/dashboard.py
   

5. Run the API:

   uvicorn scripts.api:app --reload
   

Dataset Details Name: Online Retail

Source: Online Retail dataset

Description: The dataset contains transaction data for a UK-based online retailer, including details such as InvoiceNo, StockCode, Description, Quantity, InvoiceDate, UnitPrice, CustomerID, and Country.

Future Work Extend feature engineering for improved predictive performance.

Incorporate more advanced recommendation algorithms.

Deploy the API and dashboard on cloud platforms for real-time analytics.

Instructions:

1. Create the README.md file in your project root and paste the above content.
2. Replace placeholder text (e.g., YOUR_USERNAME) with your actual GitHub username.

---

Happy analyzing!

Final Notes

• Testing: Before pushing to GitHub, run each script locally to ensure everything works as expected.
• Commit Often: Make small commits for each new feature or file.
• Documentation: Comment your code and document your project thoroughly to help others understand your work.

By following these detailed steps, you will have an impressive, well-organized data analytics project hosted on GitHub that demonstrates your skills as a data professional. Happy coding and best of luck with your project!