search

hpi-sam / rl-4-self-repair

Reinforcement Learning Models for Online Learning of Self-Repair and Self-Optimization
MIT License
0 stars 1 forks source link

Cost of swapping #4

Closed MrBanhBao closed 4 years ago

MrBanhBao commented 4 years ago

How should the cost/reward of swapping component should be implemented?

MrBanhBao commented 4 years ago

The idea of swapping components was abandoned due to the cost of possible permutations and action space changes. Component-Failure-Pairs now can be directly repaired without being on top of a stack.