Nicola Ancona

Papers

Total Citations

H-Index

About

Nicola Ancona is a researcher whose work bridges computer vision and machine learning, with a particular focus on dynamic systems and reinforcement learning. His early contributions include a foundational paper on obstacle detection using optical flow (1992, 26 citations), which demonstrated a fast, vision-based method for navigation—a precursor to modern autonomous systems. More recently, Ancona has pioneered the use of symbolic regression (SR) to construct parsimonious analytic models for dynamic systems, as highlighted in his 2020 work (25 citations). This approach addresses a critical challenge in reinforcement learning (RL): while RL excels at controlling systems with unknown dynamics, it becomes faster and safer when augmented with learned models. By employing SR to generate interpretable, compact models, Ancona’s research enables more efficient and reliable RL applications, particularly in environments with time-varying dynamics. His 2019 paper (4 citations) further explores this synergy, proposing SR as a tool to build analytic models that enhance RL performance. Ancona’s work stands out for its emphasis on model simplicity and interpretability, offering a practical bridge between classical control theory and modern learning-based methods.