Sales Prediction Analysis

Overview

This project analyzes the relationship between advertising spending across different media channels (TV, Radio, Newspaper) and sales performance. It implements multiple linear regression models to predict sales based on advertising expenditure.

Project Structure

├── data_analysis.py      # Data loading and preprocessing
├── model_training.py     # Model training and evaluation
├── visualisation.py      # Visualization functions
├── main.py              # Main execution script
├── requirements.txt      # Project dependencies
└── SalesPrediction.ipynb # Jupyter notebook with analysis

Key Features

• Data preprocessing and cleaning
• Multiple linear regression models:
  • Basic model (all features)
  • Normalized model
  • Subset model (Radio + Newspaper)
• Model performance comparison
• Visualization of results
• Feature importance analysis

Results

• TV advertising shows highest correlation with sales
• Full model achieves R² score of 0.902
• Feature importance ranking:
  1. Radio (0.107)
  2. TV (0.054)
  3. Newspaper (0.0003)

Setup and Usage

1. Install dependencies:

pip install -r requirements.txt

2. Run the analysis:

python main.py

Requirements

• Python 3.7+
• pandas
• numpy
• scikit-learn
• matplotlib
• seaborn
• openpyxl
• jupyterlab