The problem of population genetics can be viewed as a stochastic process. The aim of the project is to understand the reverse transition dynamics of the system conditioned on the end position. In order to understand the feasibility of using reinforcement learning to the problem, agents are tested against stochastic processes that resemble population genetics. Practical evidence is validated to provide a sanity check on the feasibility of deployment of such a method in practice in large-scale problems of population genetics.


Arjun Karuvally


  • Simon Gravel