Fitting Distributions to Data, A Comparison of Two Methods.
Abstract
Monte Carlo methods are used to compare the methods of maximum likelihood and least squares to estimate a cumulative distribution function. When the probabilistic model used is correct or nearly correct, the two methods produce similar results with the MLE usually slightly superior. When an incorrect model is used, or when the data is contaminated, the least squares technique often gives substantially superior results. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 30, 1981
- Accession Number
- ADA102683
Entities
People
- Paul N. Somerville
- Steven J. Bean
Organizations
- University of Central Florida