|
Registro Completo |
|
Biblioteca(s): |
Embrapa Mandioca e Fruticultura. |
|
Data corrente: |
03/04/2001 |
|
Data da última atualização: |
03/04/2001 |
|
Autoria: |
FISHER, R. A. |
|
Título: |
The precision of discriminant functions |
|
Ano de publicação: |
1940 |
|
Fonte/Imprenta: |
Ann. Eugenics., v.10, p.422-429, 1940. |
|
Idioma: |
Inglês |
|
Conteúdo: |
In a paper (1938a) on "The statistical utilization of multiple measurements" the author considered the general procedure of the establishment of discriminant functions, or sets of scores, based on an analysis for covariance, for a bactery of different experimental determinations. In general, these functions are those giving stationary to the ratio of apportionment of sums of squares between n, chosen and n2 residual degrees of freedom. In the simplest application n1 is unity, and, as was first shown by Hotelling, the primary test of measurements available are effective in making any significant discrimation, is exactly reducible to a simple analysis of variane in which p-1 degrees of freedom (for p variates) are transferred from the residual. In the general case the underlying problem of distribuition has also now been solved(Fisher, 1939). |
|
Categoria do assunto: |
-- |
|
Marc: |
LEADER 01155naa a2200121 a 4500
001 1649946
005 2001-04-03
008 1940 bl --- 0-- u #d
100 1 $aFISHER, R. A.
245 $aThe precision of discriminant functions
260 $c1940
520 $aIn a paper (1938a) on "The statistical utilization of multiple measurements" the author considered the general procedure of the establishment of discriminant functions, or sets of scores, based on an analysis for covariance, for a bactery of different experimental determinations. In general, these functions are those giving stationary to the ratio of apportionment of sums of squares between n, chosen and n2 residual degrees of freedom. In the simplest application n1 is unity, and, as was first shown by Hotelling, the primary test of measurements available are effective in making any significant discrimation, is exactly reducible to a simple analysis of variane in which p-1 degrees of freedom (for p variates) are transferred from the residual. In the general case the underlying problem of distribuition has also now been solved(Fisher, 1939).
773 $tAnn. Eugenics.$gv.10, p.422-429, 1940.
Download
Esconder MarcMostrar Marc Completo |
|
Registro original: |
Embrapa Mandioca e Fruticultura (CNPMF) |
|
