Demand Forecasting for a Retail Store

Overview

This project demonstrates demand forecasting for retail products using the "Historical Product Demand" dataset. The workflow covers data cleaning, aggregation, visualization, ARIMA modeling, and forecast evaluation.

Dataset

• Source: Kaggle - Product Demand Forecasting
• File: Historical Product Demand.csv
• Fields:
  • Product_Code
  • Warehouse
  • Product_Category
  • Date
  • Order_Demand

How to Run

1. Clone or download this repository.
2. Open demand_forecasting.ipynb in Jupyter Notebook or VS Code.
3. Run all cells in order.
   Required libraries: pandas, numpy, matplotlib, statsmodels.

To install dependencies: pip install pandas numpy matplotlib statsmodels

Project Steps

1. Data Loading: Read and preview the dataset.
2. Data Cleaning: Handle missing values and convert data types.
3. Aggregation: Aggregate demand by date and resample to weekly totals.
4. Visualization: Plot trends in daily and weekly demand.
5. Forecasting: Fit an ARIMA model to the training data.
6. Evaluation: Forecast and compare to actual demand for the test period.

Results

• The ARIMA model provides a baseline forecast for weekly demand.
• The model captures overall trends but may miss sudden spikes or drops.
• Further improvements could include using advanced models or more granular features.

Author

• Nitin Nandan
• Internship Project for CodeClause