逢甲學報, Band 13逢甲大學, 1980 |
Im Buch
Ergebnisse 1-3 von 9
Seite
... risk matrix ( 6 ) . Suppose , whether for computational simplicity or other reasons , that consideration is confined to linear estimator of the form ( 5 ) . A naive approach to the estimation proplem wonld be to proceed on the ...
... risk matrix ( 6 ) . Suppose , whether for computational simplicity or other reasons , that consideration is confined to linear estimator of the form ( 5 ) . A naive approach to the estimation proplem wonld be to proceed on the ...
Seite
... risk matrix depends on B. Without knowing B , there- fore , it is impossible to find A and b so as to minimize Risk B. The naive approach has therefore led to dilemma . Obviously , if B were known then the solution would be to set b = B ...
... risk matrix depends on B. Without knowing B , there- fore , it is impossible to find A and b so as to minimize Risk B. The naive approach has therefore led to dilemma . Obviously , if B were known then the solution would be to set b = B ...
Seite
... Risk B = AVA ' = Var B in ( 11 ) , the problem that remains is to minimize AVA ' subject to AX - I . This is known as the Gauss - Markoff Criterion . Hence , a " best " ( or minimum varia nce ) unbia- sed linear estimator of B in the ...
... Risk B = AVA ' = Var B in ( 11 ) , the problem that remains is to minimize AVA ' subject to AX - I . This is known as the Gauss - Markoff Criterion . Hence , a " best " ( or minimum varia nce ) unbia- sed linear estimator of B in the ...