Auto-association by multilayer perceptrons and singular value decomposition
The multilayer perceptron, when working in auto-association mode, is sometimes considered as an interesting candidate to perform data compression or dimensionality reduction of the feature space in information processing applications. The present paper shows that, for auto-association, the nonlinearities of the hidden units are useless and that the optimal parameter values can be derived directly by purely linear techniques relying on singular value decomposition and low rank matrix approximation, similar in spirit to the well-known Karhunen-Loève transform. This approach appears thus as an efficient alternative to the general error back-propagation algorithm commonly used for training multilayer perceptrons. Moreover, it also gives a clear interpretation of the role of the different parameters.
- Training multilayer perceptrons by principal component analysis
- Reducing the Dimensionality of Data with Neural Networks
- Dimensionality-reduction using connectionist networks
- Multilayer perceptrons and data compression
- Connection between multilayer perceptrons and regression using independent component analysis
- scientific article; zbMATH DE number 3473182 (Why is no real title available?)
- scientific article; zbMATH DE number 3525015 (Why is no real title available?)
- scientific article; zbMATH DE number 3892457 (Why is no real title available?)
- Least squares, singular values and matrix approximations
- Updating the singular value decomposition
- A comparison of three methods for principal component analysis of fuzzy interval data
- Boosting neural network feature extraction by reduced accuracy activation functions
- Feedforward neural networks for principal components extraction.
- A CNN-based shock detection method in flow visualization
- Unsupervised learning in neural computation.
- Edge-backpropagation for noisy logo recognition
- Solvable models of layered neural networks based on their differential structure
- Complex-valued autoencoders
- Processing simple geometric attributes with autoencoders
- Thermodynamics of restricted Boltzmann machines and related learning dynamics
- Non-iterative and fast deep learning: multilayer extreme learning machines
- Autoencoders reloaded
- Preserving differential privacy in deep neural networks with relevance-based adaptive noise imposition
- Sim2vec: node similarity preserving network embedding
- Reduced-order deep learning for flow dynamics. The interplay between deep learning and model reduction
- Component models for fuzzy data
- Autoencoder asset pricing models
- Deep learning with periodic features and applications in particle physics
- Multilayer perceptrons and data compression
- Dimensionality-reduction using connectionist networks
- scientific article; zbMATH DE number 6982332 (Why is no real title available?)
- scientific article; zbMATH DE number 7625173 (Why is no real title available?)
- Forward Stepwise Deep Autoencoder-Based Monotone Nonlinear Dimensionality Reduction Methods
- Linearly constrained nonsmooth optimization for training autoencoders
- Principal component analysis of ensemble recordings reveals cell assemblies at high temporal resolution
- Multilevel Artificial Neural Network Training for Spatially Correlated Learning
- Justifying and Generalizing Contrastive Divergence
- Training multilayer perceptrons by principal component analysis
- Surrogate modeling for high dimensional uncertainty propagation via deep kernel polynomial chaos expansion
- Deep nonparametric estimation of intrinsic data structures by chart autoencoders: generalization error and robustness
- An Introduction to Neural Data Compression
- Deep learning in computational mechanics: a review
- The dynamics of representation learning in shallow, non-linear autoencoders
- Risk factor aggregation and stress testing
- Functional autoencoder for smoothing and representation learning
- Enhancing the performance of variational quantum classifiers with hybrid autoencoders
- High-dimensional asymptotics of denoising autoencoders
- Auto-associative models and generalized principal component analysis
- A paired autoencoder framework for inverse problems via Bayes risk minimization
- Spectral function space learning and numerical linear algebra networks for solving linear inverse problems
- A power method for computing singular value decomposition
- Belted and Ensembled Neural Network for Linear and Nonlinear Sufficient Dimension Reduction
- Optimal linear baseline models for scientific machine learning
- A comprehensive deep learning-based approach to reduced order modeling of nonlinear time-dependent parametrized PDEs
- A comparative investigation on model selection in independent factor analysis
This page was built for publication: Auto-association by multilayer perceptrons and singular value decomposition
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1106762)
