SuperKart Sales Revenue Forecasting Model
This repository hosts a predictive machine learning model developed to forecast future sales revenue for individual products across the multi-outlet supermarket chain, SuperKart.
📋 Project Overview
SuperKart handles a wide array of retail products across diverse store tiers and geographical regions. Due to fluctuating regional consumer demands, predicting pipeline visibility and maintaining optimal stock levels is a major operational challenge. This model addresses those challenges by providing data-driven sales predictions to prevent product stockouts and minimize expensive inventory holding costs.
🛠️ Intended Uses & Limitations
- Primary Use Case: Forecasting upcoming sales revenue based on store configurations and item attributes.
- Intended Users: Supply chain managers, inventory planners, and regional retail strategists.
- Limitations: The model's predictions are highly reliant on historical tracking trends and may require recalibration if major structural market changes or unmapped seasonal shifts occur.
📊 Dataset & Feature Breakdown
The underlying model utilizes key retail variables to generate its predictive outputs:
- Product Attributes: Item Weight, Maximum Retail Price (MRP), Item Fat Content, and Visibility/Shelf Space metrics.
- Store Attributes: Outlet Establishment Year (Store Age), Outlet Size, Location Type (Tier 1, 2, or 3 cities), and Outlet Type (Supermarket vs. Grocery Store).
🚀 How to Load and Use the Model
If you saved your model as a scikit-learn pickle file, you can load and run inferences locally using the following Python snippet:
python
import joblib import pandas as pd
1. Download/load the model file from Hugging Face
model = joblib.load("your_model_filename.pkl")
2. Define sample store and product features
sample_data = pd.DataFrame([{ 'Item_Weight': 12.5, 'Item_Visibility': 0.05, 'Item_MRP': 140.0, 'Outlet_Age': 15, 'Outlet_Size_Medium': 1, 'Outlet_Type_Supermarket_Type1': 1 }])
3. Predict sales revenue
predicted_sales = model.predict(sample_data) print(f"Predicted Outlet Sales: ${predicted_sales[0]:.2f}")
Use code with caution.