Skip to main content

Table 1 Average prediction performance of the Rounds 1 and 2 of our method on test sets of the 10-fold CV.

From: Modeling DNA affinity landscape through two-round support vector regression with weighted degree kernels

Test Performance of Round 1: WD with s = 1 & m = 1
d Runtime RMSE Pearson Cor Spearman Cor
2 572 20.06 0.74 0.46
3 1034 19.99 0.74 0.47
4 1448 19.87 0.75 0.48
5 1834 19.79 0.75 0.49
6 2221 19.77 0.75 0.49
7 2430 19.76 0.75 0.50
8 2908 19.75 0.75 0.50
9 3193 19.74 0.75 0.50
Test Performance of Round 2: WD with s = 0 & m = 0
d Runtime RMSE Pearson Cor Spearman Cor
2 47 18.82 0.78 0.55
3 90 18.09 0.80 0.59
4 128 17.65 0.81 0.62
5 166 17.34 0.82 0.65
6 200 17.09 0.83 0.66
7 235 16.89 0.84 0.68
8 268 16.89 0.84 0.65
9 302 16.87 0.84 0.65
  1. Round 1 uses all k-mers up to length d, with shift = 1 and mismatch = 1. Round 2 uses only selected k-mers from Round 1, with shift = 0 and mismatch = 0. 'Runtime' includes both training and testing, in seconds. The values for the parameters selected on training data are in bold.