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

Anomaly Detection Using an Ensemble of Feature Models.

Keith Noto1, Carla Brodley, Donna Slonim

  • 1Department of Computer Science, Tufts University Medford, MA, 02155 United States.

Proceedings. IEEE International Conference on Data Mining
|October 25, 2011
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

From Noise to Knowledge: Diffusion Probabilistic Model-Based Neural Inference of Gene Regulatory Networks.

Journal of computational biology : a journal of computational molecular cell biology·2024
Same author

hnRNPL expression dynamics in the embryo and placenta.

Gene expression patterns : GEP·2023
Same author

Accurate genome-wide phasing from IBD data.

BMC bioinformatics·2022
Same author

Intergenerational effects of preconception opioids on glucose homeostasis and hepatic transcription in adult male rats.

Scientific reports·2022
Same author

Ancestry inference using reference labeled clusters of haplotypes.

BMC bioinformatics·2021
Same author

A Prospective Analysis of Genetic Variants Associated with Human Lifespan.

G3 (Bethesda, Md.)·2019
Same journal

GeoToken: Hierarchical Geolocalization of Images via Next Token Prediction.

Proceedings. IEEE International Conference on Data Mining·2026
Same journal

A Universal Metric of Dataset Similarity for Cross-silo Federated Learning.

Proceedings. IEEE International Conference on Data Mining·2026
Same journal

Heterogeneous Treatment Effect Estimation with Subpopulation Identification for Personalized Medicine in Opioid Use Disorder.

Proceedings. IEEE International Conference on Data Mining·2024
Same journal

Enhancing Personalized Healthcare via Capturing Disease Severity, Interaction, and Progression.

Proceedings. IEEE International Conference on Data Mining·2024
Same journal

Robust Unsupervised Domain Adaptation from A Corrupted Source.

Proceedings. IEEE International Conference on Data Mining·2023
Same journal

RoS-KD: A Robust Stochastic Knowledge Distillation Approach for Noisy Medical Imaging.

Proceedings. IEEE International Conference on Data Mining·2023
See all related articles

This study introduces a novel semi-supervised anomaly detection method that predicts feature values to identify outliers. Experimental results show significant performance improvements over existing techniques on diverse datasets.

Area of Science:

  • Machine Learning
  • Data Mining
  • Artificial Intelligence

Background:

  • Anomaly detection aims to identify data points deviating from a norm.
  • Traditional methods often rely on feature space distances or densities.
  • Some datasets lack clear feature space positions for normal data, hindering traditional approaches.

Purpose of the Study:

  • To develop a novel semi-supervised anomaly detection approach.
  • To address limitations of traditional methods in identifying anomalies with complex feature relationships.
  • To improve the accuracy and robustness of anomaly detection systems.

Main Methods:

  • Learning to predict feature values from other features within a training set.
  • Employing an ensemble of predictors for anomaly detection.

Related Experiment Videos

  • Utilizing a novel information-theoretic anomaly measure to combine predictor contributions and mitigate noisy features.

    • Main Results:

    • The proposed method significantly improves performance on 47 diverse datasets.
    • Outperforms current state-of-the-art feature space distance and density-based approaches.
    • The information-theoretic measure effectively handles noisy and irrelevant features.

    Conclusions:

    • The new prediction-based approach offers a powerful alternative for semi-supervised anomaly detection.
    • This method is particularly effective for datasets where anomalies have inconsistent feature relationships.
    • The approach demonstrates superior performance and robustness compared to existing methods.