Development of a model selection method for population pharmacokinetics analysis by deep-learning based reinforcement learning

The Food and Drug Administration (FDA) is seeking proposals for the development of a model selection method for population pharmacokinetics (popPK) analysis using deep-learning based reinforcement learning. The goal of this project is to create a method that can select the most appropriate model for popPK analysis, which is crucial for generic drug development. The selected method should be able to handle various scenarios of pharmacokinetic (PK) data and provide unbiased and mathematically meaningful performance metrics. The proposal should include a thorough survey of current state-of-the-art deep-learning based reinforcement learning algorithms and their applications, as well as a plan for designing simulations, identifying performance metrics, and conducting performance evaluation. • The proposal should include a detailed plan for developing a model selection method using deep-learning based reinforcement learning. • The proposal should include a plan for designing simulations that reflect different scenarios of PK data. • The proposal should include a plan for identifying proper metrics for performance evaluation. • The proposal should include a plan for conducting performance evaluation using the developed model selection method and other approaches. • The proposal should include a plan for using real PK dataset(s) to demonstrate the applicability and advantage of the developed method.