Displaying 2 results from an estimated 2 matches for "methodftrees".
2009 Mar 16
0
the effect of blocking on the size of confidence intervals - analysis using lme and lmer
...andom = ~ 1 | target, data = d)
> m2.lme = lme( response ~ method -1, random = ~ 1 | query , data = d)
Using lmer the calls are:
> m1.lmer = lmer( response ~ method -1 + ( 1 | target ), data = d)
...
> summary.lm( m0.lm )
...
Estimate Std. Error t value Pr(>|t|)
methodFTREES 0.34163 0.03221 10.61 <2e-16 ***
methodMACCS 0.45525 0.03221 14.13 <2e-16 ***
methodXXX.pg.trained 0.49498 0.03221 15.37 <2e-16 ***
methodXXX.rpg.trained 0.45049 0.03221 13.99 <2e-16 ***
...
Residual standard error: 0.2952 on 332 degre...
2009 Mar 16
0
the effect of blocking on the size of confidence intervals - analysis using aov
...uot; I found a
description, how confidence intervals of fitted parameters can be
derived from the corresponding standard errors (p. 154).
For the m0.aov model the approach is straight forward:
> summary.lm(m0.aov)
...
Coefficients:
Estimate Std. Error t value Pr(>|t|)
methodFTREES 0.34163 0.03221 10.61 <2e-16 ***
methodMACCS 0.45525 0.03221 14.13 <2e-16 ***
methodXXX.pg.trained 0.49498 0.03221 15.37 <2e-16 ***
methodXXX.rpg.trained 0.45049 0.03221 13.99 <2e-16 ***
...
Residual standard error: 0.2952 on 332 degre...