Yumi Kondo

Yumi Kondo

3 packages on CRAN

lmeNB

cran
99.99th

Percentile

The functions in this package implement the safety monitoring procedures proposed in the paper titled "Detection of unusual increases in MRI lesion counts in individual multiple sclerosis patients" by Zhao, Y., Li, D.K.B., Petkau, A.J., Riddehough, A., Traboulsee, A., published in Journal of the American Statistical Association in 2013. The procedure first models longitudinally collected count variables with a negative binomial mixed-effect regression model. To account for the correlation among repeated measures from the same patient, the model has subject-specific random intercept, which can be modelled with a gamma or log-normal distributions. One can also choose the semi-parametric option which does not assume any distribution for the random effect. These mixed-effect models could be useful beyond the application of the safety monitoring. The maximum likelihood methods are used to estimate the unknown fixed effect parameters of the model. Based on the fitted model, the personalized activity index is computed for each patient. Lastly, this package is companion to R package lmeNBBayes, which contains the functions to compute the Personalized Activity Index in Bayesian framework.

lmeNBBayes

cran
99.99th

Percentile

The functions in this package implement the safety monitoring procedures proposed in the paper titled "A flexible mixed effect negative binomial regression model for detecting unusual increases in MRI lesion counts in individual multiple sclerosis patients" by Kondo, Y., Zhao, Y. and Petkau, A.J. The procedure first models longitudinally collected count variables with a negative binomial mixed-effect regression model. To account for the correlation among repeated measures from the same patient, the model has subject-specific random intercept, which is modelled with the infinite mixture of Beta distributions, very flexible distribution that theoretically allows any form. The package also has the option of a single beta distribution for random effects. These mixed-effect models could be useful beyond the application of the safety monitoring. The inference is based on MCMC samples and this package contains a Gibbs sampler to sample from the posterior distribution of the negative binomial mixed-effect regression model. Based on the fitted model, the personalized activity index is computed for each patient. Lastly, this package is companion to R package lmeNB, which contains the functions to compute the Personalized Activity Index in the frequentist framework.

RSKC

cran
99.99th

Percentile

This RSKC package contains a function RSKC which runs the robust sparse K-means clustering algorithm.