Advanced Search

Question: maximum likelihood estimation

Question:maximum likelihood estimation

JohnZed 32 Maple
statistics
February 24 2007
0
Does anyone know how to perform maximum likelihood estimation in maple 10 or 11 with multiple unknown parameters? If pop100 is a vector of floats, then code like: MLE(LogNormal(mu, sigma), pop100) yields this error message: Error, (in Statistics:-MaximumLikelihoodEstimate) when calling '`one of {op, module () local OptionsTab; export FisherInformation, Information, Likelihood, LikelihoodRatioStatistic, LogLikelihood, MaximumLikelihoodEstimate, Score, GetValue, ProcessOptions; LogLikelihood := module () local OptionsTab; export ProcessOptions, MakeProc, GetValue; OptionsTab := table([(OptionParms)=(table([(samplesize)=[deduce, {posint, identical(deduce)}],(weights)=[Array(1..0, []), {rtable}],(ignore)=[false, {truefalse}]])),(PositionalParms)=(table([]))]); ProcessOptions := proc (parms::list, other::list, opts::table) return :-ProcessOptions(0,other,Statistics:. This happens in both Maple 10 and Maple 11, although it will work if I specify a numerical value for either mu or sigma. All of the examples in the docs have only a single unknown parameter, which seems like a very specialized case in practice... Anybody have tips on how to work around this? Thanks!
Please Wait...