Sparse reward for reinforcement learning-based continuous integration testing.
Yang YangZheng LiYing ShangQianyu LiPublished in: J. Softw. Evol. Process. (2023)
Keyphrases
- reinforcement learning
- integration testing
- black box
- object oriented programs
- action space
- state space
- model free
- continuous state and action spaces
- reinforcement learning algorithms
- software testing
- function approximation
- eligibility traces
- component based software
- test cases
- optimal policy
- reward function
- high dimensional
- temporal difference
- partially observable environments
- markov decision processes
- sparse representation
- multi agent
- machine learning
- training data
- state action
- policy gradient
- learning algorithm
- case study
- data sets
- high level
- learning agent
- learning process