Pareto Conditioned Networks
This addresses the challenge of efficiently recovering all optimal policies in multi-objective reinforcement learning, offering a scalable solution for problems with many objectives.
The paper tackles the problem of learning all Pareto-efficient policies in multi-objective optimization, which is computationally expensive due to exponential growth with objectives, by proposing Pareto Conditioned Networks (PCN) that use a single neural network to encompass all non-dominated policies, transforming optimization into classification and enabling stable, scalable learning with minimal assumptions on the Pareto front.
In multi-objective optimization, learning all the policies that reach Pareto-efficient solutions is an expensive process. The set of optimal policies can grow exponentially with the number of objectives, and recovering all solutions requires an exhaustive exploration of the entire state space. We propose Pareto Conditioned Networks (PCN), a method that uses a single neural network to encompass all non-dominated policies. PCN associates every past transition with its episode's return. It trains the network such that, when conditioned on this same return, it should reenact said transition. In doing so we transform the optimization problem into a classification problem. We recover a concrete policy by conditioning the network on the desired Pareto-efficient solution. Our method is stable as it learns in a supervised fashion, thus avoiding moving target issues. Moreover, by using a single network, PCN scales efficiently with the number of objectives. Finally, it makes minimal assumptions on the shape of the Pareto front, which makes it suitable to a wider range of problems than previous state-of-the-art multi-objective reinforcement learning algorithms.