Shortcomings of Top-Down Randomization-Based Sanity Checks for Evaluations of Deep Neural Network Explanations.
Alexander BinderLeander WeberSebastian LapuschkinGrégoire MontavonKlaus-Robert MüllerWojciech SamekPublished in: CVPR (2023)
Keyphrases
- neural network
- privacy preserving
- back propagation
- pattern recognition
- artificial neural networks
- neural network model
- genetic algorithm
- deep learning
- activation function
- learning vector quantization
- auto associative
- training algorithm
- fuzzy artmap
- associative memory
- explanatory power
- self organizing maps
- fault diagnosis
- fuzzy logic
- information systems
- neural network is trained
- generating explanations
- image reconstruction from projections
- hybrid neural network
- privacy preserving association rule mining
- feed forward neural networks
- fuzzy neural network
- causal models
- network architecture
- training process
- feed forward
- radial basis function
- high level
- artificial intelligence
- machine learning
- data mining