IPO: Interior-point Policy Optimization under Constraints
It addresses decision-making problems in real-world scenarios where constraints must be satisfied, offering a novel method for constrained RL.
The paper tackles the problem of reinforcement learning with cumulative constraints by proposing Interior-point Policy Optimization (IPO), a first-order method using logarithmic barriers, which outperforms state-of-the-art baselines in reward maximization and constraint satisfaction.
In this paper, we study reinforcement learning (RL) algorithms to solve real-world decision problems with the objective of maximizing the long-term reward as well as satisfying cumulative constraints. We propose a novel first-order policy optimization method, Interior-point Policy Optimization (IPO), which augments the objective with logarithmic barrier functions, inspired by the interior-point method. Our proposed method is easy to implement with performance guarantees and can handle general types of cumulative multiconstraint settings. We conduct extensive evaluations to compare our approach with state-of-the-art baselines. Our algorithm outperforms the baseline algorithms, in terms of reward maximization and constraint satisfaction.