A graph-based interpretability method for deep neural networks.
Tao WangXiangwei ZhengLifeng ZhangZhen CuiChunyan XuPublished in: Neurocomputing (2023)
Keyphrases
- neural network
- high accuracy
- segmentation method
- clustering method
- computational cost
- experimental evaluation
- support vector machine
- high precision
- fuzzy logic
- cost function
- pattern recognition
- detection method
- training process
- objective function
- dynamic programming
- preprocessing
- data sets
- significant improvement
- pairwise
- synthetic data
- classification accuracy
- classification method
- multiscale
- optimization method
- neural network model
- neural network is trained