This project aims to introduce the basic concepts of machine learning by implementing a simple linear regression model trained with a gradient descent algorithm to predict the price of a car based on its mileage.

In this project, i writed two programs:

• Prediction Program: This program will predict the price of a car for a given mileage. It will prompt the user for a mileage input and return the estimated price using the linear regression model.

• Training Program: This program will train the linear regression model using a dataset file containing mileage and corresponding prices. Once trained, it will save the parameters θ0 and θ1 for use in the prediction program.

The linear regression model used for prediction is defined by the following hypothesis:

estimatePrice(mileage) = θ0 + (θ1 * mileage)

Where:

• θ0 and θ1 are parameters to be learned.
• mileage is the input feature (in this case, the mileage of the car).

To train the model, we will use the gradient descent algorithm with the following formulas:

tmpθ0 = learningRate * (1/m) * Σ (estimatePrice(mileage[i]) - price[i]) tmpθ1 = learningRate * (1/m) * Σ (estimatePrice(mileage[i]) - price[i]) * mileage[i]

Where:

• learningRate is a hyperparameter determining the step size of the gradient descent.
• m is the number of training examples.
• Σ denotes the summation over all training examples.
• estimatePrice(mileage[i]) is the predicted price for the ith mileage.
• price[i] is the actual price for the ith mileage.

Ensure that your dataset file is properly formatted with mileage and price columns.

To get started with the project, follow these steps:

• Clone https://github.com/lava/matplotlib-cpp for visualization, and change the name of clonned repo to matplotlibCpp
• Clone the repository.
• make

ft_linear_regression$ make
Making MATRIX
Done
Making linear_regression_train
Done
Making linear_regression_predict
Done

• Run the training program with your dataset file to train the model and obtain θ0 and θ1.

./linear_regression_train data.csv 

• Then run the predict program

./linear_regression_predict
Enter a mileage (type 'STOP' to exit): 5000
Estimated price: 8391.82
Enter a mileage (type 'STOP' to exit): 4582
Estimated price: 8400.79
Enter a mileage (type 'STOP' to exit): 32200
Estimated price: 7808.56
Enter a mileage (type 'STOP' to exit): 100000
Estimated price: 6354.69
Enter a mileage (type 'STOP' to exit): 200000
Estimated price: 4210.34
Enter a mileage (type 'STOP' to exit): STOP
Exiting the program.