Explaining the data or explaining a model? Shapley values that uncover non-linear dependencies.
Daniel Vidali FryerInga StrümkeHien D. NguyenPublished in: CoRR (2020)
Keyphrases
- probabilistic model
- statistical model
- experimental data
- input data
- temporal dependencies
- data sets
- computational model
- data collection
- measured data
- simulation data
- mathematical model
- test data
- high quality
- high level
- synthetic data
- parameter values
- training data
- network structure
- empirical data
- data processing
- numerical data
- data samples
- prior knowledge
- neural network model
- learning models
- attribute values
- dependency graph
- relational models
- predictive model
- prior information
- raw data
- temporal information
- missing data
- image data
- data points
- end users
- data analysis
- bayesian networks
- similarity measure
- feature selection
- databases