Classification on MNIST
0.57Error RateLadder Networks
Evaluation results
| Method | Links | |
|---|---|---|
| Ladder Networks2015.11 | 0.57 | |
| VAT2015.11 | 0.64 | |
| Adversarial Autoencoderstraining=end-to-end2015.11 | 0.85 | |
| CatGAN2015.11 | 0.91 | |
| VAE (M1 + M2)2015.11 | 0.96 | |
| NN BaselineRegularization=Dropout2015.11 | 1.25 | |
| BNBatch Normalization=true2015.10 | 52 | |
| uSF+BNFeedback=Uniform Sign-concordant, Batch Normalization=true2015.10 | 55 | |
| NuSFFeedback=Normalized uSF2015.10 | 60 | |
| SGDOptimizer=SGD2015.10 | 67 | |
| BN+BMBatch Normalization=true, Batch Manhattan=true2015.10 | 73 | |
| brSF+BN+BMFeedback=Biased Random Sign-concordant, Batch Normalization=true, Batch Manhattan=true2015.10 | 80 | |
| uSF+BN+BMFeedback=Uniform Sign-concordant, Batch Normalization=true, Batch Manhattan=true2015.10 | 83 | |
| frSF+BN+BMFeedback=Fixed Random Sign-concordant, Batch Normalization=true, Batch Manhattan=true2015.10 | 91 | |
| LeNet-N# Trained Params=64k2020.03 | 92 | |
| uSF+BMFeedback=Uniform Sign-concordant, Batch Manhattan=true2015.10 | 95 | |
| BMBatch Manhattan=true2015.10 | 99 | |
| uSFFeedback=Uniform Sign-concordant2015.10 | 8,865 |