Using mechanistic machine learning to forecast glucose and infer physiologic phenotypes in the ICU: what is possible and what are the challenges.
David J. AlbersMatthew E. LevineAndrew M. StuartJan ClaassenBruce J. GluckmanGeorge HripcsakPublished in: AMIA (2018)
Keyphrases
- machine learning
- intensive care unit
- lessons learned
- decision trees
- real world
- inductive learning
- knowledge acquisition
- learning algorithm
- information extraction
- data mining
- text classification
- learning tasks
- feature selection
- computational biology
- machine learning methods
- clinical practice
- computational intelligence
- decision support system
- support vector machine
- knowledge discovery
- reinforcement learning
- natural language processing
- long term
- key issues
- active learning
- high risk
- artificial intelligence
- caenorhabditis elegans