This project implements a retail sales forecasting and monitoring system focused on the US retail market. It provides real-time analysis and predictive insights into retail sales performance, helping stakeholders make data-driven decisions in the retail trade sector.
The platform analyzes total sales value generated by retail establishments in the United States, covering both durable and non-durable goods sectors. This comprehensive analysis provides valuable insights into:
- Consumer spending behavior patterns
- Overall retail industry health
- Sales trends and seasonal variations
- Future sales projections with confidence intervals
- Live monitoring of retail sales performance
- Interactive data visualization
- Trend analysis and pattern recognition
- Seasonal variation tracking
- Two-year sales forecasting capability
- Confidence interval calculations
- Seasonal pattern recognition
- Trend analysis and decomposition
The project implements several forecasting models, with the Holt-Winters model selected as the optimal solution.
- Source: Federal Reserve Economic Data (FRED)
- Time Range: January 1992 to April 2023
- Forecast Period: May 2023 to April 2025
- Data Type: Advance Retail Sales (Retail Trade Sector)
- Clone the repository
git clone https://github.com/JP3132/Data-Driven-Retail-Sales-Analytics-Platform.git
cd Data-Driven-Retail-Sales-Analytics-Platform- Set up virtual environment (Recommended)
python -m venv env
# For Windows
env\Scripts\activate.bat
# For Unix or MacOS
source env/bin/activate- Install dependencies
pip install -r requirements.txt- Run the application
streamlit run app.py- Access the web interface
http://localhost:8501- Inventory Management: Optimize stock levels based on forecasted demand
- Resource Planning: Efficient allocation of resources based on predicted sales
- Strategic Decision Making: Data-driven insights for business strategy
- Risk Management: Understanding uncertainty through confidence intervals