Dream Team project
- Thierry Bossy
- Raphael Strebel
- Ignacio S. Aleman
Project 1 of the "Machine Learning" course (CS433), given in the Fall semester of 2019.
To run our project, create a folder named "data" at the root of the project directory and put the test.csv and train.csv files in it.
Navigate to "src" and run python run.py. (Note that you need at least Python 3.6 to run the code)
Methods in implementations.py use different cost functions or gradient computation that can be found in cost.py. A big generic gradient descent method is implemented and used by all gradient descent models with different parameters.
Definition of gradient_descent method:
def gradient_descent(y, tx, compute_loss, compute_gradient, initial_w, max_iters=0, gamma=10e-6, batch_size=None, num_batch=None, debugger=None, dynamic_gamma=False):