Commonly used sequential decision making tasks such as the games in the Arcade Learning Environment (ALE) provide rich observation spaces suitable for deep reinforcement learning. However, they consist mostly of low-level control tasks which are of limited use for the development of explainable artificial intelligence(XAI) due to the fine temporal resolution of the tasks. Many of these domains also lack built-in high level abstractions and symbols. Existing tasks that provide for both strategic decision-making and rich observation spaces are either difficult to simulate or are intractable. We provide a set of new strategic decision-making tasks specialized for the development and evaluation of explainable AI methods, built as constrained mini-games within the StarCraft II Learning Environment.

No abstract

No abstract

Common approaches to learn complex tasks in reinforcement learning include reward shaping, environmental hints, or a curriculum. Yet few studies examine how they compare to each other, when one might prefer one approach, or how they may complement each other. As a first step in this direction, we compare reward shaping, hints, and curricula for a Deep RL agent in the game of Minecraft. We seek to answer whether reward shaping, visual hints, or the curricula have the most impact on performance, which we measure as the time to reach the target, the distance from the target, the cumulative reward, or the number of actions taken. Our analyses show that performance is most impacted by the curriculum used and visual hints; shaping had less impact. For similar navigation tasks, the results suggest that designing an effective curriculum and providing appropriate hints most improve the performance. Common approaches to learn complex tasks in reinforcement learning include reward shaping, environmental hints, or a curriculum, yet few studies examine how they compare to each other. We compare these approaches for a Deep RL agent in the game of Minecraft and show performance is most impacted by the curriculum used and visual hints; shaping had less impact. For similar navigation tasks, this suggests that designing an effective curriculum with hints most improve the performance.

Patricia Chaffey has had a passion for studying and designing interaction between humans and technology since her undergraduate career at Mount Holyoke College, and continues to pursue this interest at the University of Southern California. Some of her notable work includes developing a robotic learning companion and designing a simulation to study how people interact with swarms of robots using a virtual agent as an intermediary. Patricia has received awards to support her travel to conferences and leadership workshops, which include, but are not limited to, the 2018 ELIS Expanding Horizons award, and the 2017 Computing Research Association -Women Grace Hopper Celebration Research Scholar award. Patricia has participated in a number of publications across the different labs she has been active in. When not in the lab, Patricia enjoys volunteering with BOTS (Building Opportunities with Teachers in Schools), where she works with elementary teachers and their students on robotics.