Max Robotics — Global Service Robot Directory + US Certification

About

Alexandra Kearney is a machine learning researcher whose work sits at the intersection of reinforcement learning, human-machine interaction, and adaptive algorithm design. Her research has made meaningful contributions to the field of temporal-difference (TD) learning, with a particular focus on making these methods more robust and autonomous through intelligent parameter adaptation. Kearney's most notable contributions include the development of TIDBD, an algorithm that automatically adapts step-sizes in TD learning — a longstanding challenge in the field — removing the burden of manual hyperparameter tuning that has historically limited practical deployment. Building on this, her work on learning feature relevance through step-size adaptation advances the broader goal of meta-learning as a tool for representation learning, enabling algorithms to identify which inputs matter most during training. Perhaps her most applied contribution explores the use of reinforcement learning to support human decision-making in prosthetic limb control, demonstrating her commitment to translating algorithmic advances into real-world assistive technology. With citations spanning multiple years and research threads, Kearney's work reflects a coherent research vision: making learning systems smarter, more self-sufficient, and ultimately more useful to the humans who depend on them.