Rapid development of information technology, together with advances in sensory and data acquisition techniques, brings increasing necessity to handle datasets from various sources. In recent years, transfer learning has emerged as an effective framework to combine information from multiple domains and tackle related tasks in target domains by transferring previous-acquired knowledge from source domains. In transfer learning, statistical models and methodologies are widely involved and play a critical role, which, however, is not particularly emphasized in literature. In this talk, I will review transfer learning techniques with a focus on statistical models and statistical methodologies, demonstrating how statistics helps to conduct transfer learning in various scenarios. After that, I will highlight some research opportunities to extend transfer learning applications to statistical process control.