Generalized SHAP: Generating multiple types of explanations in machine learning.
Dillon BowenLyle UngarPublished in: CoRR (2020)
Keyphrases
- multiple types
- machine learning
- networked data
- co occurrence
- heterogeneous networks
- learning algorithm
- decision trees
- machine learning algorithms
- real world
- class labels
- active learning
- supervised learning
- natural language processing
- multiple features
- data sets
- support vector machine
- heterogeneous information networks
- data mining
- feature selection
- learning tasks
- contextual information
- semi supervised learning
- visual information
- feature extraction
- reinforcement learning
- prior knowledge
- unsupervised learning
- higher level
- text classification
- mobile devices
- information extraction
- low level