The following homework focuses on a semi-supervised learning task. This type of task includes both labeled and unlabeled samples. The goal is to assign the correct label to unlabeled samples. This kind of problem is very close to real-world scenarios where you have a lot of data but few labels (unlabeled data). The paradigm of reference for the resolution of the task was "similar features share similar labels".