Bayesian Classification and Regression Trees for Predicting Incidence of Cryptosporidiosis
2011

Using Bayesian Models to Predict Cryptosporidiosis Incidence

Sample size: 1332 publication 10 minutes Evidence: moderate

Author Information

Author(s): Hu Wenbiao, O'Leary Rebecca A., Mengersen Kerrie, Low Choy Samantha

Primary Institution: Queensland University of Technology

Hypothesis

Can Bayesian classification and regression trees effectively model the incidence of cryptosporidiosis in Queensland, Australia?

Conclusion

The Bayesian CART model is effective for identifying and estimating the spatial distribution of cryptosporidiosis risk.

Supporting Evidence

  • The Bayesian CART model allowed for flexible identification of nonlinear interactions between variables.
  • The study found that temperature significantly influenced cryptosporidiosis incidence.
  • Bayesian methods provided better predictive accuracy compared to traditional frequentist methods.

Takeaway

This study used a special kind of math model to understand how weather affects a sickness called cryptosporidiosis, which can make people sick.

Methodology

The study compared Bayesian CART models with Bayesian spatial CAR models to analyze cryptosporidiosis incidence data.

Potential Biases

Potential biases may arise from the selection of variables and the model structure.

Limitations

The dataset had a large number of zero incidence rates, which may affect the model's performance.

Participant Demographics

The study analyzed cryptosporidiosis cases from local government areas in Queensland, Australia.

Statistical Information

P-Value

0.1046

Confidence Interval

1.02–1.21

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.1371/journal.pone.0023903

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication