Figure 3

Performance comparison of various ranges of k. On the x-axis, ‘5’ means that k ranged from 1 to 5. Training data comprised 22,389 human RefSeq mRNA transcripts and 22,389 GENCODE lncRNA transcripts. SVM classifiers were trained using 10-fold cross-validation on the training datasets. The figure indicates that the computation load rises and the accuracy increases along as k increases.