Prediction with some naive Bayes classifiers for circular data
Prediction with some naive Bayes classifiers for circular data.
vmnb.pred(xnew, mu, kappa, ni) spmlnb.pred(xnew, mu1, mu2, ni)
xnew |
A numerical matrix with new predictor variables whose group is to be predicted. Each column refers to an angular variable. |
mu |
A matrix with the mean vectors expressed in radians. |
mu1 |
A matrix with the first set of mean vectors. |
mu2 |
A matrix with the second set of mean vectors. |
kappa |
A matrix with the kappa parameters for the vonMises distribution or with the norm of the mean vectors for the circular angular Gaussian distribution. |
ni |
The sample size of each group in the dataset. |
Each column is supposed to contain angular measurements. Thus, for each column a von Mises distribution or an circular angular Gaussian distribution is fitted. The product of the densities is the joint multivariate distribution.
A numerical vector with 1, 2, ... denoting the predicted group.
Michail Tsagris
R implementation and documentation: Michail Tsagris mtsagris@uoc.gr.
x <- matrix( runif( 100, 0, 1 ), ncol = 2 ) ina <- rbinom(50, 1, 0.5) + 1 a <- vm.nb(x, x, ina) a2 <- vmnb.pred(x, a$mu, a$kappa, a$ni)
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.