A training period without foreground objects is not always available in some environments, and this makes bootstrapping the background model hard.
Ninety-five perconfidencedence intervals were estimated by bootstrapping, and the model was selected using a likelihood ratio test.
Second, we bootstrapped the full model 2,000 times and obtained bias-corrected and accelerated (BCa) 95% confidence intervals to assess the stability of the regression coefficients from logistic regression.
The 95% confidence intervals of excess rates were estimated by bootstrapping the residuals of the full model 1,000 times.
A 10,000-replication bootstrap of the model was then performed on the sample, and 95% bootstrap confidence intervals were collected.
In most instances in which there was a difference >5% points in the bootstrap values, the model with more parameters produced higher support values (10 of 13 instances; Fig. 5).
From this bootstrap sample, the model (4) is estimated by maximizing (5) in which the simulated parameters β ^ b e ∗ are used instead of β ^ e. β ^ b r is the resulting estimation of the relative regression coefficients.
Analysed with the GEE bootstrapped models, the effect estimate was identical (IRR: 0.63) to that in the negative binomial regression model, albeit with a wider confidence interval (95% CI: 0.40 0.99).
Firstly, the datasets informing each of the extrapolation regression models (10 cost models and 3 survival models) were bootstrapped, and the corresponding model coefficients re-estimated to generate new estimates of lifetime costs and survival for each eligible patient.
Bootstrapping demonstrated that the model slope was accurate and significantly different from zero (Mean = 0.513; quantiles0.05,0.95 = 0.248, 0.894).
