Using Shapley Values and Variational Autoencoders to Explain Predictive Models with Dependent Mixed Features.
Lars H. B. OlsenIngrid Kristine GladMartin JullumKjersti AasPublished in: J. Mach. Learn. Res. (2022)
Keyphrases
- classification models
- statistical models
- predictive power
- feature extraction
- feature values
- prior knowledge
- co occurrence
- predictive model
- complex systems
- model selection
- feature set
- denoising
- low level
- predictive modeling
- predictive clustering trees
- abstraction levels
- learning algorithm
- historical data
- learning models
- game theory
- optical flow
- feature space
- image segmentation