Online greedy identification of linear dynamical systems
This work addresses efficient exploration for linear dynamical systems, but it is incremental as it builds on existing experimental design frameworks.
The authors tackled the problem of exploration in unknown linear dynamical systems by introducing an online greedy policy that maximizes information gain per step, resulting in experimentally competitive performance with low complexity compared to gradient-based methods.
This work addresses the problem of exploration in an unknown environment. For linear dynamical systems, we use an experimental design framework and introduce an online greedy policy where the control maximizes the information of the next step. In a setting with a limited number of experimental trials, our algorithm has low complexity and shows experimentally competitive performances compared to more elaborate gradient-based methods.