In recent years, robust Markov decision processes (MDPs) have emerged as a prominent modeling framework for dynamic decision problems affected by uncertainty.

Using this information, we would like to solve the optimization problem (1).

The latter allows us to control the error in the stochastic gradient updates due to inaccurate solution to both subproblems.

A natural idea is to compute the solution path w.r.t.

Our main contribution is a generic framework, called Social-Inverse, for handling migrations between tasks of diffusion enhancement and diffusion containment. For Social-Inverse, we present theoretical analysis to obtain insights regarding how different contagion management tasks can be subtly correlated in order for samples from one task to help the optimization of another task.