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

One-bit-matching conjecture for independent component analysis.

Zhi-Yong Liu1, Kai-Chun Chiu, Lei Xu

  • 1Department of Computer Science and Engineering, Chinese University of Hong Kong, Shatin, New Territories. zyliu@cse.cuhk.edu.hk

Neural Computation
|March 10, 2004
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

[Survey on the occupational musculoskeletal disorder and its risk factors among male steelworkers].

Zhonghua yu fang yi xue za zhi [Chinese journal of preventive medicine]·2013
Same author

Characterisation and identification of dihydroindole-type alkaloids from processed semen strychni by high-performance liquid chromatography coupled with electrospray ionisation ion trap time-of-flight mass spectrometry.

Phytochemical analysis : PCA·2013
Same author

Identification and effect decomposition of risk factors for Brucella contamination of raw whole milk in china.

PloS one·2013
Same author

Different toxicity of the novel Bacillus thuringiensis (Bacillales: Bacillaceae) strain LLP29 against Aedes albopictus and Culex quinquefasciatus (Diptera: Culicidae).

Journal of economic entomology·2013
Same author

Study on the growth and the photosynthetic characteristics of low energy C(+) ion implantation on peanut.

PloS one·2013
Same author

Proteomic analysis reveals that proteasome subunit beta 6 is involved in hypoxia-induced pulmonary vascular remodeling in rats.

PloS one·2013
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

This study provides the first mathematical proof for the one-bit-matching conjecture in independent component analysis (ICA), confirming source separation under specific conditions. Experiments also show robustness when assumptions are relaxed.

Area of Science:

  • Signal Processing
  • Machine Learning
  • Statistical Analysis

Background:

  • The one-bit-matching conjecture in Independent Component Analysis (ICA) posits that sources can be separated if kurtosis signs match between source and model probability density functions (PDFs).
  • This conjecture is widely believed and supported by various ICA algorithms but lacks theoretical mathematical proof.
  • Existing ICA methods implicitly rely on this conjecture for successful source separation.

Purpose of the Study:

  • To provide a rigorous mathematical proof for the one-bit-matching conjecture in ICA.
  • To investigate the conjecture's robustness when the assumption of vanishing skewness in model densities is violated.
  • To demonstrate that kurtosis maximization is a specific instance of the minimum mutual information criterion.

Main Methods:

Related Experiment Videos

  • Mathematical derivation focusing on skewness and kurtosis of probability density functions (PDFs).
  • Theoretical proof developed under the assumption of vanishing skewness for model densities.
  • Empirical experiments conducted to assess the conjecture's robustness against violations of the vanishing skewness assumption.

Main Results:

  • A mathematical proof is established for the one-bit-matching conjecture in ICA, contingent on vanishing skewness of model densities.
  • Empirical evidence demonstrates the conjecture's resilience even when the vanishing skewness assumption is not strictly met.
  • It is shown that the kurtosis maximization criterion is a special case of the minimum mutual information criterion.

Conclusions:

  • The one-bit-matching conjecture for ICA is theoretically validated under the specified conditions.
  • The findings support the practical applicability of ICA algorithms relying on this conjecture.
  • This work clarifies the relationship between different ICA optimization criteria.