Interpretable pipelines with evolutionary optimized modules for reinforcement learning tasks with visual inputs

Abstract

The importance of explainability in AI has become a pressing concern, for which several explainable AI (XAI) approaches have been recently proposed. However, most of the available XAI techniques are post-hoc methods, which however may be only partially reliable, as they do not reflect exactly the state of the original models. Thus, a more direct way for achieving XAI is through interpretable (also called glass-box) models. These models have been shown to obtain comparable (and, in some cases, better) performance with respect to black-boxes models in various tasks such as classification and reinforcement learning. However, they struggle when working with raw data, especially when the input dimensionality increases and the raw inputs alone do not give valuable insights on the decision-making process. Here, we propose to use end-to-end pipelines composed of multiple interpretable models co-optimized by means of evolutionary algorithms, that allows us to decompose the decision-making process into two parts: computing high-level features from raw data, and reasoning on the extracted high-level features. We test our approach in reinforcement learning environments from the Atari benchmark, where we obtain comparable results (with respect to black-box approaches) in settings without stochastic frame-skipping, while performance degrades in frameskipping settings.

The intersection of evolutionary computation and explainable AI

“…Similar attempts at combining RL and EC have tried to obtain interpretable policies for RL tasks by combining decision trees induced by GP or Grammatical Evolution with RL acting on the leaves while the policy interacts with the environment [14][15][16]29]. Some other works in this area have explicitly focused on addressing the interpretability question in white-box models.…”

Interpretable AI for policy-making in pandemics

“…In [5], the authors use genetic programming [9] and CMA-ES [7] to evolve interpretable DTs that are able to work in RL settings with images as input. However, the experimental results show that the proposed approach exhibits good performance only in scenarios that are not affected by noise.…”