Interpretable Convolutional Neural Networks via Feedforward Design
By C.-C. Jay Kuo, Min Zhang, Siyang Li, Jiali Duan, Yueru Chen. 2018
177 citations recorded by the source corpus.
Abstract The model parameters of convolutional neural networks (CNNs) are determined by backpropagation (BP). In this work, we propose an interpretable feedforward (FF) design without any BP. The FF design adopts a data-centric approach. It derives network parameters of the current layer based on data statistics from the output of the previous layer in a one-pass manner. To construct convolutional layers, we develop a new signal transform, called the Saab (Subspace approximation with adjusted bias) transform. It is a variant of the principal component analysis with an added bias vector to annihilate activation’s nonlinearity. Multiple Saab transforms in cascade yield multiple convolutional layers. As to fully-connected layers, we construct them using a cascade of multi-stage linear least squared regressors. The classification and robustness performances of BP- and FF-designed CNNs applied to the MNIST and the CIFAR-10 datasets are compared. Finally, we comment on the relationship between BP and FF designs.
Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Jiali Duan
- Yueru Chen
- Chau Yuen
- Yuren Zhou
- Meng Zhang
- Billy Pik Lik Lau
- U-Xuan Tan
- Marakkalage Sumudu Hasala
In Computer Vision and Pattern Recognition
Citations in the corpus
- A survey of data fusion in smart city applications (2019, methodology): “Alternate approach has proposed to use subspace approximation with an adjusted bias technique [260] to build interpretable CNN, which uses feed forward design to better explain the model’s choice in allocating certain hyper-parameters.”
- An Interpretable Generative Model for Handwritten Digits Synthesis (2019, methodology): “By following [29], we decompose input vectors into DC (direct current) and AC (alternating current) two components.”