CDT: Cascading Decision Trees for Explainable Reinforcement Learning.
Zihan DingPablo Hernandez-LealGavin Weiguang DingChangjian LiRuitong HuangPublished in: CoRR (2020)
Keyphrases
- decision trees
- reinforcement learning
- machine learning
- optimal policy
- function approximation
- naive bayes
- machine learning algorithms
- decision tree algorithm
- decision tree induction
- state space
- predictive accuracy
- decision rules
- markov decision processes
- learning algorithm
- training data
- random forest
- dynamic programming
- multi agent
- rule induction
- robotic control
- attribute selection
- feature construction
- reinforcement learning algorithms
- temporal difference
- learning process
- optimal control
- ensemble methods
- data sets
- classification rules
- bayesian networks
- action space
- delaunay triangulation
- constructive induction
- genetic algorithm
- temporal difference learning
- autonomous learning
- policy search
- supervised learning