The system is built on a retail sales dataset containing product attributes such as Item Weight, Item Visibility, Item MRP (price), Item Type, and Outlet details. These features are used to classify products into High, Medium, or Low sales categories, helping businesses understand product performance.
Demo Streamlit link:https://salesprediction-42qh5fyfwswlyoubnr6zw5.streamlit.app/
- app.py → Main Streamlit application file that runs the web interface.
- requirements.txt → List of required Python libraries for installation.
- sales_classifier.joblib → Pre-trained machine learning models, scaler, and feature information (loaded during execution).
- data/ (optional) → Folder for storing input CSV files for batch prediction.
- Clone the repository:
git clone <your-repo-link> cd <your-repo-name>
- Install dependencies:
pip install -r requirements.txt
- Run the application:
streamlit run app.py
- Access the system in your browser at
http://localhost:8501.
The system loads AI models (SVM, KNN, ANN) from a saved .joblib file.
Users can:
- View a homepage to know about the features of the modules, comparision of default models and tuned models.
- Enter product details in the sidebar for single predictions, to predit whether it is low, medium or high.
- Upload a CSV file for batch predictions, and can download the result.
The model processes input features, applies data preprocessing (encoding & scaling), and predicts the sales category. The results are displayed in an interactive interface, with options to download predictions as a CSV.