Oded Maron

Brown University, Massachusetts Institute of Technology

Papers

Total Citations

121

H-Index

About

Oded Maron is a researcher whose work lies at the intersection of machine learning, probabilistic modeling, and robotics. He is best known for his pioneering contributions to learning finite automata with stochastic output functions, a framework that elegantly handles uncertainty in sequential decision-making tasks. His 1995 paper on this topic, which has accumulated over 50 citations, introduced a novel approach to inferring hidden structures from noisy data, with a particularly creative application to map learning—enabling robots to build spatial representations from imperfect sensor readings. Maron also made significant strides in model selection with his 1997 work on the Racing algorithm, a computationally efficient method for comparing multiple learning algorithms or parameter settings. This work, cited over 40 times, offered a principled way to allocate evaluation resources, making it especially valuable for lazy learning systems where speed and accuracy are critical. Though his publication record is compact, Maron’s ideas have had enduring influence, particularly in robotics and adaptive systems, where his stochastic automata and racing methods continue to inspire new generations of researchers tackling problems in autonomous navigation and efficient model selection.