We then used a software package called Genes@Work for differential expression discovery, and SVM light linear kernel for building classification models.
In a Monte Carlo simulation within the AFNI software package, using a smoothing kernel of 8.08 mm, a connection radius of 7.72 mm on 3.48 mm × 3.48 mm × 6 mm voxels, it was determined that an activation volume of 1,278 μl satisfied the P < 0.001 threshold.
The statistical model was calculated with help of the ksvm function from the kernlab package in R statistical software by using a polynomial kernel matrix similarity (degree = 2), a cost parameter C = 1 and a 13-fold cross validation (leave-ne-out cross validation).
In a Monte Carlo simulation within the AFNI software package and using a smoothing kernel of 8 mm and a connection radius of 7.2 mm on 3.44 × 3.44 × 6 mm voxels, it was determined that an activation volume of 102 original voxels (2754 microliters) satisfied the P < 0.01 threshold.
The spot enumeration was done with a kernel density estimation based algorithm [15] using a Gaussian kernel.
In brief, the software counts the read coverage per base along the genome and calculates a signal value by smoothing these counts using a Tukey kernel.
Mutual information was computed using a Gaussian Kernel estimate.
Local cell density was estimated with a kernel density estimator using a Gaussian kernel on eight different scales.
The spatial distribution of cows and heifers was mapped using a kernel density function [ 21] of the same software (Fig. 1).
Multiple kernel learning (MKL) aims at learning a combination of different kernels in order to better match the underlying problem instead of using a single fixed kernel.
