Sub-Optimality is All that's Crucial

I am currently researching on the theoretical aspects of Transfer Reinforcment Learning with my team and advising from Prof. Sichun Gao in UCSD's Computer Science & Engineering Department. We are developing a conceptual framework that may be used for families of algorithms and trying to build an understanding of how sub-optimality conditions may be more suitable for conducting transfer learning particularly through the probabilisti inference lens of controls inference.