Image Classification on CIFAR-10 8-way Non-IID
78.9Test AccuracyEnsemble
Evaluation results
| Method | Links | |
|---|---|---|
| EnsembleBackbone=VGG112025.06 | 78.9 | |
| LPBackbone=VGG11, Weight Initialization=initialized with the weights of the first base model2025.06 | 73.8 | |
| Individual ModelsBackbone=VGG112025.06 | 72.3 | |
| KF-Linear DeepLIFTBackbone=VGG11, Neuron Importance Score=DeepLIFT2025.06 | 71.6 | |
| KF-Linear ConductanceBackbone=VGG11, Neuron Importance Score=Conductance2025.06 | 71.4 | |
| KF-Linear UniformBackbone=VGG11, Neuron Importance Score=Uniform2025.06 | 69.5 | |
| KF-Gradient ConductanceBackbone=VGG11, Neuron Importance Score=Conductance2025.06 | 69.4 | |
| KF-Gradient DeepLIFTBackbone=VGG11, Neuron Importance Score=DeepLIFT2025.06 | 69 | |
| KF-Gradient UniformBackbone=VGG11, Neuron Importance Score=Uniform2025.06 | 67 | |
| KDBackbone=VGG11, Weight Initialization=randomly initialized2025.06 | 42.4 | |
| OTF DeepLIFTBackbone=VGG11, Neuron Importance Score=DeepLIFT2025.06 | 13.4 | |
| OTF UniformBackbone=VGG11, Neuron Importance Score=Uniform2025.06 | 11.3 | |
| OTF ConductanceBackbone=VGG11, Neuron Importance Score=Conductance2025.06 | 11 | |
| Vanilla AveragingBackbone=VGG112025.06 | 10 |