Final Project: Linear Regression Analysis of the "Auto-mpg" Dataset

Authors

• Omar A. Mohammed
• Omar M. Ahmed
• Amgad A. Shaban
• Mahmoud M. Abdelaty
  Department of Systems and Biomedical Engineering, Faculty of Engineering, Cairo University

Overview

This project conducts a linear regression analysis on the Auto-mpg dataset to investigate the relationship between various car attributes and fuel efficiency (miles per gallon, mpg). The analysis identifies key factors such as cylinders, displacement, horsepower, weight, and model year that impact fuel efficiency, with a focus on predicting mpg using these variables. The project addresses outliers and assumptions, while providing suggestions for future research.

Key Features

• Dataset: Auto-mpg dataset from Kaggle, containing car attributes from the late 1970s and early 1980s.
• Techniques: Linear regression analysis, data cleaning, feature engineering, and outlier detection.
• Key Variables: mpg, cylinders, displacement, horsepower, weight, acceleration, model year, origin.
• Data Visualization: Scatter plots, bar plots, box plots for correlation analysis and outlier identification.
• Outlier Handling: Applied Kolmogorov-Smirnov test and IQR methods for accurate outlier removal.
• Model: Developed linear regression model with Ordinary Least Squares (OLS) for prediction.

Results

The study reveals significant predictors of fuel efficiency, providing valuable insights for vehicle manufacturers, policymakers, and consumers in optimizing fuel consumption.

Requirements

• Python 3.x
• Libraries: pandas, matplotlib, seaborn, numpy, scipy, scikit-learn

Installation

1. Clone the repository:

   git clone https://github.com/yourusername/auto-mpg-linear-regression.git