This may not be very relevant to Competitive Programming but somehow related.

I am working on the following problem. For each student, we have a list of the courses he can take in the upcoming semester (he has already finished the prerequisites, and those courses must be part of the degree plan). From each department, we have all the information needed to make the schedule like how many faculty members are available, how many sections they can offer, how many sections in parallel can be offered, and the types of courses (lectures, labs, combined, ... ).

The objective is to schedule the classes and give each student a customized schedule based on the available data. I want to know if someone has worked on a similar thing before. One thing I thought of is to find clusters in the data and solve for those groups independently. But I am not sure if that can be accomplished using Machine learning algorithms.

Any suggestion or help will be appreciated.

Thanks