Statements (22)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
gptkb:convolutional_neural_network
Unsupervised learning algorithm |
| gptkbp:application |
Feature extraction
Dimensionality reduction Image denoising Pretraining deep networks |
| gptkbp:hasComponent |
gptkb:Decoder
Encoder |
| https://www.w3.org/2000/01/rdf-schema#label |
Denoising autoencoder
|
| gptkbp:input |
Corrupted data
|
| gptkbp:lossFunction |
Reconstruction loss
|
| gptkbp:output |
Clean data
|
| gptkbp:proposedBy |
gptkb:Pascal_Vincent
2008 |
| gptkbp:relatedTo |
Boltzmann machine
Noise reduction Representation learning |
| gptkbp:trainedTo |
Reconstruct input from corrupted version
|
| gptkbp:usedIn |
Machine learning
Deep learning |
| gptkbp:bfsParent |
gptkb:Boltzmann_machine
|
| gptkbp:bfsLayer |
5
|