Extracting and composing robust features with denoising autoencoders
Pascal Vincent, Hugo Larochelle, Yoshua Bengio et al. · International Conference on Machine Learning · 2008
Previous work has shown that the difficulties in learning deep generative or discriminative models can be overcome by an initial unsupervised learning step that maps inputs to useful intermediate representations. We introduce and motivate a new training principle for unsupervised learning of a representation based on the idea of making the learned representations robust to partial corruption of the input pattern. This approach can be used to train autoencoders, and these denoising autoencoders can be stacked to initialize deep architectures. The algorithm can be motivated from a manifold learning and information theoretic perspective or from a generative model perspective. Comparative experiments clearly show the surprising advantage of corrupting the input of autoencoders on a pattern classification benchmark suite. 1.
Cited by 7,191 works. doi:10.1145/1390156.1390294
Authors
References
- Learning representations by back-propagating errors
- Neural networks and physical systems with emergent collective computational abilities.
- A Fast Learning Algorithm for Deep Belief Nets
- Reducing the Dimensionality of Data with Neural Networks
- Greedy Layer-Wise Training of Deep Networks
- Learning Deep Architectures for AI
- Scaling learning algorithms towards AI
- Efficient Learning of Sparse Representations with an Energy-Based Model
- An empirical evaluation of deep architectures on problems with many factors of variation
- Image Denoising Via Sparse and Redundant Representations Over Learned Dictionaries