Backprop-MPDM: Faster Risk-Aware Policy Evaluation Through Efficient Gradient Optimization

@article{Mehta2018BackpropMPDMFR,
  title={Backprop-MPDM: Faster Risk-Aware Policy Evaluation Through Efficient Gradient Optimization},
  author={Dhanvin Mehta and Gonzalo Ferrer and Edwin Olson},
  journal={2018 IEEE International Conference on Robotics and Automation (ICRA)},
  year={2018},
  pages={1740-1746}
}
In Multi-Policy Decision-Making (MPDM), many computationally-expensive forward simulations are performed in order to predict the performance of a set of candidate policies. In risk-aware formulations of MPDM, only the worst outcomes affect the decision making process, and efficiently finding these influential outcomes becomes the core challenge. Recently, stochastic gradient optimization algorithms, using a heuristic function, were shown to be significantly superior to random sampling. In this… CONTINUE READING

References

Publications referenced by this paper.
Showing 1-10 of 20 references

Fast discovery of influential outcomes for risk-aware MPDM

2017 IEEE International Conference on Robotics and Automation (ICRA) • 2017
View 8 Excerpts
Highly Influenced

Socially aware motion planning with deep reinforcement learning

2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) • 2017
View 1 Excerpt

Autonomous navigation in dynamic social environments using Multi-Policy Decision Making

2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) • 2016
View 3 Excerpts

Planning for robotic exploration based on forward simulation

Robotics and Autonomous Systems • 2016
View 1 Excerpt

Intention-aware online POMDP planning for autonomous driving in a crowd

2015 IEEE International Conference on Robotics and Automation (ICRA) • 2015
View 1 Excerpt