This is a problem where we wish to predict the attempts a user will make to solve a given problem. We have been given information about the users, Problems and attempts users made for problems. We are trying to solve this by a very powerful yet simple approach called Matrix factorization. A more advanced and effiecient implementation can be applied using the state of the art implementation such as GraphLab's factorization_recommender but even this simple appraoch has provided very good results and I have managed to get at 7th position in AV's recommendation practice problem.