Nathan Sprague

Kalamazoo College, James Madison University

Papers

Total Citations

H-Index

About

Nathan Sprague’s research bridges the critical gap between high-dimensional machine learning and human-robot interaction, with a focus on developing autonomous systems that are both intelligent and trustworthy. His foundational work, “Predictive projections” (2009, 16 citations), introduced a novel linear dimensionality reduction algorithm that enables learning control policies in extremely high-dimensional state spaces. By discovering projections where simple nearest-neighbor methods can accurately predict future states, Sprague provided an elegant solution to a core challenge in reinforcement learning and robotics, laying groundwork for more efficient autonomous decision-making. More recently, his study “Convergence Across Behavioral and Self-report Measures Evaluating Individuals' Trust in an Autonomous Golf Cart” (2022, 6 citations) addresses the human side of automation. This work rigorously validates that self-reported trust measures align with actual behavioral responses, offering a reliable framework for assessing user trust in autonomous systems—a critical factor for adoption in military, healthcare, and consumer applications. Sprague’s dual focus on algorithmic efficiency and human-centered evaluation marks him as a thoughtful contributor to the future of safe, accepted autonomous technologies.