Abram L. Friesen

Papers

Total Citations

H-Index

About

Abram L. Friesen is a researcher whose work bridges the critical intersection of continuous optimization and developmental robotics. His primary research areas include nonconvex optimization, robot learning by imitation, and Bayesian approaches to artificial intelligence. Friesen made a significant contribution to optimization theory with his work on "Recursive Decomposition for Nonconvex Optimization" (2016, 27 citations), which addresses the fundamental challenge that most real-world optimization problems—from vision to probabilistic inference—are nonconvex, causing standard techniques to find only local optima. This work provides a principled framework for tackling these difficult problems. In robotics, Friesen proposed an innovative "Bayesian Developmental Approach to Robotic Goal-Based Imitation Learning" (2015, 14 citations), which draws inspiration from the developmental hypothesis that children use self-experience to bootstrap the process of learning by observing others. This approach offers a principled probabilistic framework for enabling robots to learn new skills from human demonstrations. Friesen's work is notable for its elegant combination of rigorous mathematical foundations with practical applications in artificial intelligence and robotics, making important strides toward building robots that can learn more naturally and solve complex optimization problems.