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

Computational Techniques for Spatial Logistic Regression with Large Datasets.

Christopher J Paciorek1

  • 1Harvard School of Public Health, 655 Huntington Avenue, Boston, MA 02115.

Computational Statistics & Data Analysis
|April 30, 2008
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

Associations between type of blood collection, analytical approach, mean haemoglobin and anaemia prevalence in population-based surveys: A systematic review and meta-analysis.

Journal of global health·2022
Same author

National, regional, and global estimates of anaemia by severity in women and children for 2000-19: a pooled analysis of population-representative data.

The Lancet. Global health·2022
Same author

Life expectancy and risk of death in 6791 communities in England from 2002 to 2019: high-resolution spatiotemporal analysis of civil registration data.

The Lancet. Public health·2021
Same author

The forests of the midwestern United States at Euro-American settlement: Spatial and physical structure based on contemporaneous survey data.

PloS one·2021
Same author

Quantifying trends and uncertainty in prehistoric forest composition in the upper Midwestern United States.

Ecology·2019
Same author

NONLINEAR PREDICTIVE LATENT PROCESS MODELS FOR INTEGRATING SPATIO-TEMPORAL EXPOSURE DATA FROM MULTIPLE SOURCES.

The annals of applied statistics·2018
Same journal

Modeling Disease-specific Survival in Observational Studies with Missing Cause of Death: Leveraging Information from Clinical Trial Data.

Computational statistics & data analysis·2026
Same journal

A simultaneous confidence-bounded true discovery proportion perspective on localizing differences in smooth terms in regression models.

Computational statistics & data analysis·2025
Same journal

MIXANDMIX: numerical techniques for the computation of empirical spectral distributions of population mixtures.

Computational statistics & data analysis·2024
Same journal

Locally sparse quantile estimation for a partially functional interaction model.

Computational statistics & data analysis·2024
Same journal

Flexible Regularized Estimation in High-Dimensional Mixed Membership Models.

Computational statistics & data analysis·2024
Same journal

GPU Accelerated Estimation of a Shared Random Effect Joint Model for Dynamic Prediction.

Computational statistics & data analysis·2024
See all related articles

A novel Bayesian spectral basis model effectively identifies geographic risk factors for non-normal health outcomes. This spatial modeling approach offers better sensitivity and specificity than penalized likelihood methods, with good computational efficiency.

Area of Science:

  • Epidemiology
  • Spatial Statistics
  • Biostatistics

Background:

  • Epidemiological research often involves non-normal health outcomes, large sample sizes, and small effect sizes.
  • Accurate spatial modeling is crucial for linking health outcomes to geographic risk factors, especially with non-normal data.
  • Existing methods like penalized quasi-likelihood (PQL) may struggle with overfitting in spatial analyses.

Purpose of the Study:

  • To compare the performance of penalized likelihood and Bayesian spatial models for non-normal data.
  • To introduce and develop a Bayesian spectral basis model using Fourier basis for spatial analysis.
  • To evaluate model fit, speed, and ease of implementation for different spatial modeling techniques.

Main Methods:

  • Comparison of penalized likelihood models (including PQL) and Bayesian models.

Related Experiment Videos

  • Development and application of a Bayesian spectral basis model utilizing Fourier basis representation.
  • Evaluation using simulations for sensitivity, specificity, and computational efficiency.
  • Illustration on a real-world dataset of cancer cases in Taiwan.

    • Main Results:

    • The Bayesian spectral basis model demonstrated the best balance of sensitivity and specificity in simulations.
    • Penalized likelihood methods were found to be prone to overfitting.
    • A Bayesian Markov random field model was computationally efficient but statistically less performant than the spectral basis model.
    • The spectral basis model showed promise for both binary and count data in spatial modeling.

    Conclusions:

    • The Bayesian spectral basis model is a powerful and efficient tool for analyzing spatial patterns in non-normal health outcomes.
    • This spectral basis approach offers advantages over penalized likelihood methods, particularly in controlling overfitting.
    • The developed model is recommended for its performance in epidemiological studies and potential applicability to complex hierarchical models.