Advertising Sales Regression

This repository contains the source code for a Streamlit app that demonstrates Simple and Multiple Linear Regression models applied to advertising sales data. The models were created and explained in detail in the following blog posts:

• Simple Linear Regression
• Multiple Linear Regression

Table of Contents

• Overview
• Features
• Installation
• Usage
• App Demo
• Data
• Modeling
• Contributing
• License

Overview

This project showcases the application of Simple and Multiple Linear Regression models to predict sales based on advertising data. The models were built using Python and deployed using Streamlit, a popular framework for creating interactive web applications. The app allows users to explore the data, visualize relationships, and make predictions based on the regression models.

Features

• Data Visualization: Interactive plots to explore the relationship between advertising channels (TV, Radio, Newspaper) and sales.
• Model Building: Simple and Multiple Linear Regression models are built and explained.
• Prediction Interface: Users can input their own values for advertising spend to see the predicted sales.
• Explanation: The app includes brief explanations of how the models work and what the results mean.

Installation

To run this application locally, you need to have Python installed. Then, follow these steps:

1. Clone the repository:

   git clone https://github.com/surajwate/advertising-sales-regression.git
   cd advertising-sales-regression

2. Create a virtual environment (optional but recommended):

   python -m venv env
   source env/bin/activate  # On Windows use `env\Scripts\activate`

3. Install the required packages:

   pip install -r requirements.txt

4. Run the Streamlit app:

   streamlit run app.py

Usage

Once the app is running, you'll be able to interact with the regression models through the web interface:

• Input Data: Adjust the sliders to input different values for TV, Radio, and Newspaper advertising budgets.
• View Results: The app will display the predicted sales based on the regression models.
• Visualize: Explore the interactive plots to understand the relationship between the different variables.

App Demo

You can try out the live version of the app here: Advertising Sales Regression App

Data

The dataset used in this project is the Advertising Dataset, which includes data on TV, Radio, and Newspaper advertising budgets and their corresponding sales. This dataset is a classic example used to explain linear regression.

• Features:
  • TV: Advertising budget spent on TV (in thousands of dollars).
  • Radio: Advertising budget spent on Radio (in thousands of dollars).
  • Newspaper: Advertising budget spent on Newspapers (in thousands of dollars).
• Target:
  • Sales: Sales generated (in thousands of units).

Modeling

Simple Linear Regression

• Model: Predicts sales based on a single independent variable (e.g., TV advertising).
• Interpretation: The slope of the regression line indicates how much sales are expected to increase for each additional dollar spent on TV advertising.

Multiple Linear Regression

• Model: Predicts sales based on multiple independent variables (TV, Radio, and Newspaper advertising).
• Interpretation: Shows how sales are influenced by each advertising channel, accounting for the effects of the others.

Contributing

Contributions are welcome! If you have suggestions, bug reports, or improvements, feel free to open an issue or submit a pull request.

1. Fork the repository.
2. Create a new branch (git checkout -b feature-branch).
3. Make your changes.
4. Commit your changes (git commit -m 'Add new feature').
5. Push to the branch (git push origin feature-branch).
6. Open a pull request.

License

This project is licensed under the MIT License - see the LICENSE file for details.