Amiram Moshaiov
Papers
9
Total Citations
88
H-Index
6
About
Amiram Moshaiov is a leading figure in evolutionary robotics, whose work fundamentally rethinks how robots learn to navigate complex environments. His research centers on the intersection of evolutionary multi-objective optimization and neuro-controller design, addressing the critical challenge of evolving both the topology and weights of neural networks for robotic control. Moshaiov’s major contributions include pioneering the simultaneous multi-objective evolution of neuro-controllers, treating navigation as a problem of conflicting objectives like speed and straight-line motion. His work on "family bootstrapping" introduced a genetic transfer learning approach to reduce the designer knowledge needed for evolving robots on complex, related tasks, a concept that has garnered significant attention with 18 citations. With a cumulative impact of over 88 citations across his most-cited papers, Moshaiov has systematically compared optimizers like MO-CMA-ES and NSGA-II for neuro-controller evolution, and advanced the field by promoting transfer optimization through many-objective topology and weight evolution. His notable achievement includes the conceptualization of "Multi-competence Cybernetics," a framework for studying multiobjective artificial and multi-fitness natural systems, solidifying his role as an innovator in adaptive robotic intelligence.
Research Focus
Key Achievements
Top Papers
- 1Multi-objective topology and weight evolution of neuro-controllers19 citations · 2016
- 2
- 3Multi-objective evolution of robot neuro-controllers12 citations · 2009
- 4
- 5
- 6
- 7
- 8
- 9