Genetic and environmental smoothing of lactation curves with cubic splines

I.M.S. White, S. Brotherstone, R. Thompson

Research output: Contribution to journalArticlepeer-review

Abstract

Most approaches to modeling lactation curves involve parametric curves with fixed or random coefficients. In either case, the resulting models require the specification of an underlying parametric curve. The fitting of splines represents a semiparametric approach to the problem. In the context of animal breeding, cubic smoothing splines are particularly convenient because they can be incorporated into a suitably constructed mixed model. The potential for the use of splines in modeling lactation curves is explored with a simple example, and the results are compared with those using a random regression model. The spline model provides greater flexibility at the cost of additional computation. Splines are shown to be capable of picking up features of the lactation curve that are missed by the random regression model.
Original languageEnglish
Pages (from-to)632-638
Number of pages7
JournalJournal of Dairy Science
Volume82
Issue number3
Publication statusPublished - 1 Mar 1999

Fingerprint

Dive into the research topics of 'Genetic and environmental smoothing of lactation curves with cubic splines'. Together they form a unique fingerprint.

Cite this