Figure 4From: Prediction of plant pre-microRNAs and their microRNAs in genome-scale sequences using structure-sequence features and support vector machineDetermination of the best feature subset. Two indicators named LooErrorRate and TestErrorRate were used for the best subset evaluation. The LooErrorRate was calculated with 5-fold cross validation model. The TestErrorRate was calculated by independent training set and testing set with optimized parameters. The parameters of penalty coefficient c and the kernel function parameter g were obtained by grid search method.Back to article page