Firstly, in order to determine whether there is a significant overall difference in the temporal relationships between the two subgroups, a common VAR model for all subjects is tested against a VAR model with separate parameters for each group using a likelihood ratio test.
LOOCV entails training the model (fitting both the second-level ANOVA and the linear SVM model) with all subjects minus one, and using the remaining single individual to test the accuracy of the prediction.
Among males, the optimal multivariate model included 30 transcripts, including 6 that overlap with the best fitting regression model for all subjects: ASB9, H2A histone family, member J (H2AFJ), hexose-6-phosphate dehydrogenase (H6PD), IL17RB, poly(A) binding protein interacting protein 1 (Pand1), and PVALB (see Additional File 5).
We conducted further analyses restricting the data set to the 71% of the population with employment history, and adding a variable for working at the chemical plant to the model in Table 2. Overall, the R for the model for this subset was 58%, similar to the R of 55% for the model with all subjects.
For the model with all subjects, there were 26 participating subjects, resulting in 77 complete building experiments.
Typically, physiological studies on real-time detection of drowsiness usually use the same model for all subjects.
As encountered by other investigators [28], [29], we found violations of the Wing & Kristofferson model across all subjects.
Recent studies suggest that given the large potential errors in measurement of the AIF, it is better to use an assumed, fixed AIF model for all subjects [49].
The primary model included all subjects with RA and all years of follow-up.
Current smoking and ex-smoking were included when applying a model with all subjects.
