Estimation.RUM.MultiType.MLE: Performs parameter estimation for a Multitype Random Utility...

In StatRank: Statistical Rank Aggregation: Inference, Evaluation, and Visualization

Description

This function supports RUMs 1) Normal 2) Normal with fixed variance (fixed at 1) 3) Exponential

Usage

Estimation.RUM.MultiType.MLE(Data, K = 2, iter = 10, dist, ratio = 0.2,
  race = FALSE)

Arguments

Data	data in either partial or full rankings
K	number of components in mixture distribution
iter	number of EM iterations to run
dist	underlying distribution. Can be "norm", "norm.fixedvariance", "exp"
ratio	parameter in the algorithm that controls the difference of the starting points, the bigger the ratio the more the distance
race	TRUE if data is sub partial, FALSE (default) if not

Value

results from the inference

Examples

Data.Tiny <- matrix(c(1, 2, 3, 3, 2, 1, 1, 2, 3), ncol = 3, byrow = TRUE)
Estimation.RUM.MultiType.MLE(Data.Tiny, K=2, iter = 3, dist= "norm.fixedvariance")

StatRank documentation built on May 1, 2019, 8:22 p.m.