The coefficients assigned to the Laplacians and the kernels are optimized in parallel.
The κ parameter in (12) is replaced by the coefficients assigned on each individual data sources.
For the coefficients assigned informative priors, the choice of the prior plays an important role in determining the type of shrinkage of estimates of effects induced.
The main contribution of the present work lies in the additive combination of multiple kernels and Laplacians; moreover, the coefficients assigned to the kernels and the Laplacians are optimized automatically.
Reproducing precise values of selection coefficients proved a challenge, with the assumption of constant selection between sampling points leading to a systematic underestimate in the coefficients assigned to driver loci.
However, since the main contribution of this article is to optimize the combination of multiple kernels and Laplacians, the coefficients assigned on each kernel and Laplacian still need to be optimized.
Analogously, the coefficients assigned on kernels can also be obtained in the similar formulation, given by (19) where most of the variables are defined in the similar way as Q3 in (15).
We expect that the other two OKLC models (models 1 and 3) optimize the coefficients assigned on the kernels and the Laplacians to leverage multiple sources in integration and, at the same time, to increase the robustness of the combined model on combining relevant and irrelevant data sources.
Using the coefficients assigned to each predictor in the logistic regression model, we derived a prediction model as follows: logit (predicted probability of infant with EONI) = -4.047 + 0.055*CRP + 0.013*[(WBC count/10000)^7.6354 - 1.660, where EONI was coded (0) if absent and (1) if present.
