Evaluation of Machine Learning Approaches for Estimating Individualized Treatment Regimens for Time-to-Event Outcomes in Observational Studies