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

Antipredictable sequences: harder to predict than random sequences

H Zhu1, W Kinzel

  • 1The Santa Fe Institute,1399 Hyde Park Road, Santa Fe, MN 87501, USA. zhuh@santafe.edu

Neural Computation
|November 6, 1998
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

CD4+ T cell apoptosis induced by anti-CD4 antibodies.

Journal of Tongji Medical University = Tong ji yi ke da xue xue bao·2003
Same author

Screening of novel epilepsy-related genes and isolation and identification of cDNAs.

Journal of Tongji Medical University = Tong ji yi ke da xue xue bao·2003
Same author

Construction and analysis of three-dimensional graphic model of single-chain Fv derived from an anti-human placental acidic isoferritin monoclonal antibody by computer.

Journal of Tongji Medical University = Tong ji yi ke da xue xue bao·2003
Same author

Effects of gastric electrical field stimulation with long pulses on gastric emptying in dogs.

Neurogastroenterology and motility·2003
Same author

Clinical effects and mechanism of chanlibao in accelerating second stage of labor.

Journal of Tongji Medical University = Tong ji yi ke da xue xue bao·2003
Same author

Anti-proliferative effects induced by anti-CD4 human/murine chimeric antibody and murine anti-CD4 monoclonal antibody.

Journal of Tongji Medical University = Tong ji yi ke da xue xue bao·2003
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

Any prediction algorithm can be outperformed by a simpler algorithm designed to exploit its weaknesses. This highlights that effective prediction relies on assumed prior data distributions.

Area of Science:

  • Machine Learning
  • Theoretical Computer Science
  • Information Theory

Background:

  • Sequence prediction algorithms aim to forecast future data points.
  • Assessing the performance of prediction algorithms is crucial.
  • The relationship between an algorithm's complexity and its predictability is not fully understood.

Purpose of the Study:

  • To demonstrate a fundamental limitation in discrete-state sequence prediction algorithms.
  • To show that for any predictor, a counter-sequence can be constructed.
  • To illustrate the inherent reliance of effective predictors on prior assumptions.

Main Methods:

  • Theoretical analysis of prediction algorithms.
  • Construction of adversarial sequences.

Related Experiment Videos

  • Illustrative example using a simple neural network for bit sequence prediction.

    • Main Results:

    • For any prediction algorithm A, a simpler algorithm B exists to generate a sequence where A consistently fails.
    • If algorithm A performs well on sequence x, a sequence y exists where A performs equally poorly.
    • The phenomenon is general and applicable across various prediction algorithms.

    Conclusions:

    • Effective sequence prediction is not solely based on algorithmic power but heavily relies on implicit prior data distributions.
    • This finding has implications for algorithm design and performance evaluation.
    • Understanding these limitations is key to developing more robust predictive models.