Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Nonlinear autoassociation is not equivalent to PCA.

N Japkowicz1, Jos, M A Gluck

  • 1Faculty of Computer Science, Dalhousie University, Halifax, Nova Scotia, Canada, B3H 1W5.

Neural Computation
|April 19, 2000
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Novel age-dependent learning deficits in a mouse model of Alzheimer's disease: implications for translational research.

Neurobiology of aging·2009
Same author

A neurocomputational model of tonic and phasic dopamine in action selection: a comparison with cognitive deficits in Parkinson's disease.

Behavioural brain research·2009
Same author

Probabilistic categorization: how do normal participants and amnesic patients do it?

Neuroscience and biobehavioral reviews·2008
Same author

Basal ganglia and dopamine contributions to probabilistic category learning.

Neuroscience and biobehavioral reviews·2007
Same author

Risk and protective haplotypes of the alpha-synuclein gene associated with Parkinson's disease differentially affect cognitive sequence learning.

Genes, brain, and behavior·2007
Same author

Dopaminergic contribution to cognitive sequence learning.

Journal of neural transmission (Vienna, Austria : 1996)·2007
Same journal

A Model-Free Reinforcement Learning Implementation of Decision Making Under Uncertainty by Sequential Sampling.

Neural computation·2026
Same journal

DROP: Distributional and Regular Optimism and Pessimism for Reinforcement Learning.

Neural computation·2026
Same journal

Hierarchical Active Inference Using Successor Representations.

Neural computation·2026
Same journal

W-Kernel and Its Principal Space for Frequentist Evaluation of Bayesian Estimators.

Neural computation·2026
Same journal

A Hidden Markov Model-Inspired Sequence Classification Method for Hyperdimensional Computing.

Neural computation·2026
Same journal

Sparse Graphical Modeling for Electrophysiological Phase-Based Connectivity Using Circular Statistics.

Neural computation·2026
See all related articles

Nonlinear autoassociators outperform linear methods like PCA for latent extraction and classification. They learn complex domains by creating multimodal error surfaces, enabling superior nonlinear recognition capabilities.

Area of Science:

  • Machine Learning
  • Artificial Intelligence
  • Neural Networks

Background:

  • A common misconception is that autoassociators with nonlinearities are equivalent to linear methods like Principal Component Analysis (PCA).
  • This study addresses the performance differences between linear and nonlinear autoassociators.

Purpose of the Study:

  • To demonstrate that nonlinear autoassociators differ significantly from linear methods.
  • To show that nonlinear autoassociators can outperform linear methods in latent extraction, projection, and classification tasks.

Main Methods:

  • Training autoassociators with backpropagation, incorporating nonlinearities in the hidden layer.
  • Analyzing the reconstruction error surfaces generated by both linear and nonlinear autoassociators.

Related Experiment Videos

Main Results:

  • Nonlinear autoassociators exhibit multimodal error reconstruction surfaces with multiple local valleys, unlike the flat or unimodal surfaces of linear autoassociators.
  • This interpolation bias allows nonlinear autoassociators to effectively classify nonlinear multimodal domains.

Conclusions:

  • Nonlinear autoassociators are not equivalent to linear methods and offer superior performance for specific tasks.
  • Autoassociators with hidden unit nonlinearities are capable of performing nonlinear classification and recognition.