Skip to main content
Figure 3 | BMC Bioinformatics

Figure 3

From: MiRenSVM: towards better prediction of microRNA precursors using an ensemble SVM classifier with multi-loop features

Figure 3

The Architecture of miRenSVM. The original negative samples in the training set are divided into k equal partitions (k ranges from 1 to the ratio of negative samples to positive samples). The final decision is made by aggregating the results of k sub-SVM classifiers that are trained by the entire positive samples and a partition of negative samples. Two aggregation methods are considered in this work: majority vote and mean distance.

Back to article page
\