The Construction of Algorithms for the Efficient Estimation of Multivariate Probability Densities.
Abstract
Techniques have been proposed for estimating probability densities nonparmetrically. These include penalized maximum likelihood, series, fixed kernel and nearest neighbour procedures. Special attention was to the two and three dimensional cases. A databased random number generator for arbitrarily high dimensions was also developed and applied to ballistics data at Aberdeen Proving Ground. A number of biomedical applications were proposed for the related problem of the nonparametric estimation of intervals using cancer data. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 20, 1982
- Accession Number
- ADA110766
Entities
People
- James R. Thompson
Organizations
- Rice University