The Mileage Prediction project is all about forecasting vehicle fuel economy using different algorithm analysis in terms of miles per gallon (mpg). It is a modified version of the StatLib library dataset and contains features such as cylinders, displacement, horsepower, weight, acceleration, model year and origin. The raw data must be imported and preprocessed to remove missing values and scaling attributes. Data visualization helps us interpret the relations among variables. The dependent variable is mpg while independent variables include displacement, horsepower, weight and acceleration. Next we split the dataset into two parts namely training set and testing set where linear regression model will be trained on the training data. This performance of this model can be assessed with metrics like Average Absolute Error (MAE), Mean Absolute Percentage Error (MAPE) or R² Score. Finally, predictions are made based on the test data by applying these results to determine how accurately it was able to predict car performance based only upon information provided about those four quantities: displacement/weight ratio; number of cylinders/horsepower ratings; age/year made; country where manufactured/to whom sold abroad? update : more models were added into the new notebook