Over the past decade, decision trees have been widely used to represent controllers (a.k.a. policies) in an explainable way, with dtControl2 as a current state-of-the-art tool. However, for large systems or those with many corner cases, such representations tend to be too complex and not human-comprehensible. Reducing the size of the decision tree is not straightforward, as missing just a single crucial case might result in an incorrect controller. We tackle this issue in the setting of Markov decision processes by extending dtControl2 with the "\varepsilon" functionality: Given an allowed imprecision \varepsilon \geq 0, we construct a smaller decision tree, distilling the essence of the controller while still guaranteeing its \varepsilon-optimality. This enables us to provide tunably simpler explanations, omitting a controllable amount of detail. Our tool constructs decision trees that are orders of magnitude smaller than the state of the art.
Blogger's Review: This paper effectively balances explainability and controller optimality by introducing a tunable error parameter \varepsilon, offering a fresh perspective on decision tree applications, especially in complex system scenarios.