Advertising Sales Prediction using Machine Learning

Project Overview

This project predicts sales based on advertising budgets spent on TV, Radio, and Newspaper. It uses machine learning techniques to understand the relationship between advertising cost and sales performance.

---

Project Files

File Name	Description
Advertising Budget and Sales dataset.csv	Dataset containing advertising budgets and sales
Advertising_to_sales_cost_prediction.ipynb	Main notebook with complete analysis and model
README.md	Project documentation

---

Dataset Description

Column Name	Description
TV	Budget spent on TV advertising
Radio	Budget spent on Radio advertising
Newspaper	Budget spent on Newspaper advertising
Sales	Sales generated from advertising

---

Step-by-Step Process

Step	Description
Step 1	Imported required libraries (pandas, numpy, matplotlib, seaborn, sklearn)
Step 2	Loaded dataset using pandas
Step 3	Checked data types and structure
Step 4	Cleaned data (removed null values and duplicates)
Step 5	Performed exploratory data analysis
Step 6	Created visualizations (heatmap, pairplot, scatter plots)
Step 7	Selected features (TV, Radio, Newspaper) and target (Sales)
Step 8	Split data into training and testing sets
Step 9	Applied machine learning model (Linear Regression / Random Forest)
Step 10	Evaluated model performance

---

Visualizations Used

Visualization	Purpose
Heatmap	Shows correlation between variables
Pair Plot	Shows relationships between features
Scatter Plot	Shows effect of each advertising channel on sales

---

Model Information

Model	Description
Linear Regression	Used as base model for prediction
Random Forest (if used)	Handles complex patterns and improves accuracy

---

Evaluation Metrics

Metric	Description
R2 Score	Measures how well the model predicts data
MSE	Measures average squared error
RMSE	Measures prediction error in same units as sales

---

Key Findings

Insight	Explanation
TV Advertising	Strongest impact on sales
Radio Advertising	Moderate impact
Newspaper Advertising	Least impact
Combined Effect	TV and Radio together give better results

---

How to Run the Project

Step	Command / Action
1	Open Jupyter Notebook
2	Load Advertising_to_sales_cost_prediction.ipynb
3	Run all cells step by step
4	View outputs, graphs, and predictions

---

Output

Feature	Description
Sales Prediction	Predicts sales based on advertising budget
Data Insights	Shows which channel performs better
Visual Analysis	Graphs for understanding patterns

---

Conclusion

This project demonstrates how advertising budgets influence sales and how machine learning can help in making better marketing decisions.

---

Author

Name	Roll Number
Kanchan Kapri	1240258215

Happy coding

la la la la la.........