Machine learning (ML) is progressively becoming an indispensable instrument for enhancing and automating healthcare. It plays a crucial role in addressing significant challenges, such as prediction tasks and decision support, by leveraging ML techniques on electronic health records, genetic sequencing data, and clinical trial records. In this project we mainly focus on policy evaluation for sequential time series data. Within our team, we engage in machine learning research that encompasses both the development of methodology and practical applications. Consequently, we rely on training deep learning and classical statistical models. In other words, not all our projects necessitate GPU utilization but rather derive greater benefits from parallel CPU computation.