Accuracy of different network types for MNIST image classification. The accuracy of the network in this study based on the constrained DAG is superior to the case of the random DAG and is comparable to a few deep neural networks with common architecture
| Network type | Accuracy | Reference |
|---|---|---|
| Random DAG from |
99% | [7] |
| Random DAG from Watts-Strogatz model | 99.3% | [7] |
| Convolutional neural network |
99.6% | [21] |
| Constrained DAG from |
99.6% | This work |
| Deep neural network |
99.6% | [22] |
*Data augmentation was applied to the images before feeding to the networks.