A data-driven approximate dynamic programming approach based on association rule learning: Spacecraft autonomy as a case study.
Gianni D'AngeloMassimo TipaldiFrancesco PalmieriLuigi GlielmoPublished in: Inf. Sci. (2019)
Keyphrases
- rule learning
- approximate dynamic programming
- data driven
- linear program
- subgroup discovery
- dynamic programming
- inductive logic programming
- reinforcement learning
- rule sets
- decision rules
- step size
- control policy
- learning scheme
- inductive learning
- linear programming
- learning classifier systems
- background knowledge
- hill climbing
- relational learning
- cost function
- neural network
- long run
- optimal solution
- training data
- rough sets
- data mining