Advertisement Budget Sales Prediction using linear regression

📌 Description

A Linear Regression project to analyze and predict sales based on TV, Radio, and Newspaper advertisement budgets.

---

📁 Project Overview

This project demonstrates how linear regression can be applied to a marketing dataset to determine the relationship between advertising spend and sales performance.

---

🧪 Key Features

• Linear Regression using Scikit-learn
• Data visualization with Seaborn and Matplotlib
• Model evaluation using MAE, MSE, R² Score, and cross-validation
• Outlier detection and comparison of model performance before and after removing outliers

---

📊 Results Summary

Metric	With Outliers	Without Outliers
Accuracy (R²)	89.94%	87.82%
MAE	0.28	1.39
MSE	0.12	3.26

Insights:

• TV and Radio have the strongest influence on sales.
• Newspaper budget has minimal impact.
• Removing outliers reduces accuracy slightly but offers more stable predictions.

---

✅ Model Usage Guidance

• Use the one without outliers when you want a more stable and fair predictor.
• Use the one with outliers when you're okay with higher sensitivity but possible overfitting.

---

📎 How to Run

1. Clone the repository
2. Open the Jupyter/Colab notebook
3. Run all cells to view analysis and predictions