Corpus ID: 7305965

Optimal Binary Autoencoding with Pairwise Correlations

@article{Balsubramani2017OptimalBA,
  title={Optimal Binary Autoencoding with Pairwise Correlations},
  author={A. Balsubramani},
  journal={ArXiv},
  year={2017},
  volume={abs/1611.02268}
}
We formulate learning of a binary autoencoder as a biconvex optimization problem which learns from the pairwise correlations between encoded and decoded bits. Among all possible algorithms that use this information, ours finds the autoencoder that reconstructs its inputs with worst-case optimal loss. The optimal decoder is a single layer of artificial neurons, emerging entirely from the minimax loss minimization, and with weights learned by convex optimization. All this is reflected in… Expand

References

SHOWING 1-10 OF 43 REFERENCES
Extracting and composing robust features with denoising autoencoders
Autoencoders, Unsupervised Learning, and Deep Architectures
  • P. Baldi
  • Mathematics, Computer Science
  • ICML Unsupervised and Transfer Learning
  • 2012
Generative Moment Matching Networks
Generalized Denoising Auto-Encoders as Generative Models
Importance Weighted Autoencoders
Convex Deep Learning via Normalized Kernels
Provable Bounds for Learning Some Deep Representations
Auto-Encoding Variational Bayes
Breaking the Curse of Dimensionality with Convex Neural Networks
  • F. Bach
  • Computer Science, Mathematics
  • J. Mach. Learn. Res.
  • 2017
...
1
2
3
4
5
...