DeepTLE: Learning Code-Level Features to Predict Code Performance before It Runs.
Meiling ZhouJie ChenHaiyang HuJiaCheng YuZhongjin LiHua HuPublished in: APSEC (2019)
Keyphrases
- prior knowledge
- learning algorithm
- source code
- knowledge acquisition
- feature construction
- online learning
- feature extraction
- image features
- co occurrence
- learning scheme
- inductive learning
- high level abstractions
- computer programming
- levels of abstraction
- multiple features
- inductive inference
- learning tasks
- mobile learning
- low level
- active learning
- learning process
- feature space
- bayesian networks
- case study