The task of choosing the right courses for students is of utmost importance yet very difficult, very often pushing the students to inappropriate decisions that will impair academic success and motivation. At present, aging students take the advice of overly simple or general opinions, with blind-eyed preferences, to choose courses that do not fit their potentials or, in other words, their objectives. This paper offers a solution by developing a course selection prediction model based on machine learning-a model that will make course recommendations specifically tailored to each student. The model analyzes historical student data, including grades and prior course performance, to predict the students succeeding most likely in their chosen courses. The analytical-hierarchy process (AHP) of the study intends to add an additional layer to the model by fitting in other parameters such as course difficulty level, student preferences, and the quality of the faculty teaching the course. The study design gives methods to arrive at a model starting from the phase of data collection, cleaning, and teaching with algorithms like decision trees and logistic regression techniques. The main aim is to help students make wise choices in their courses, improve performance, and also assist universities in handling their course offerings better.