Relevance Which evidence is relevant for this?
We considered using just the top 10 features, where the ranking is based on the correlation coefficient, minimum redundancy maximum relevance [ 31] (which has been applied to other human studies [ 32]), mutual information (which has been used successfully in mRNA research [ 33]), using the top SVM weights, and recursive feature elimination via SVM [ 34].
In this study, we use maximum relevance minimum redundancy (MRMR), which is a frequently used feature selection algorithm.
There are many feature evaluation approaches available, and the minimum redundancy maximum relevance (mRMR) algorithm [65], which can find the optimal features with minimum redundancy, was used in this study.
In this study, we utilized the maximum relevance minimum redundancy mRMR algorithm, which reduces the features' dimensions by selecting the most relevant features while removing the redundant ones.
It sorts a feature based on score function which is maximum relevance to target and minimum redundancy to the already selected features.
Another feature selection method we used is the Minimum Redundancy - Maximum Relevance (MRMR) method proposed in [ 16], which has been proved very effective for microarray data analysis.
Signature genes were the selected differentially expressed genes in HCC metastatic vs. non-metastatic samples, with the Student t-test Benjadjustedchberg adjusted p value < 0.001, and further ranked by the method of minimum redundancy and maximum relevance(MRMR) [ 19](Additional File 1), which resulted in a list of 349 ranked candidate genes.
The remaining features were further refined using the minimum redundancy maximum relevance (mRMR) method and incremental feature selection (IFS), which are feature selection methods that have been widely used in recent years [ 34, 55– 58].
MRNET is based on the maximum relevance and minimum redundancy feature selection method [ 25], which is a significantly different procedure than the one employed by C3NET but also ARACNE, RN or CLR.
