<p>The comparison of recognition results with different block partitions by using the baseline settings: <i>R</i><sub><i>x</i></sub> = 1, <i>R</i><sub><i>y</i></sub> = 1, <i>R</i><sub><i>t</i></sub> = 3, 6<sup><i>t</i></sup><i>h</i> order polynomial kernel, LOSO cross validation.</p
<p>Each boxplot summarizes Kappa for classifying the metabolic inheritance patterns (mIPs) from 41 e...
<p>Comparison of prediction accuracy on four multiclass classification datasets by varying the numbe...
<p>For each subgraph, no significant difference in pairwise comparisons is marked with “~”, otherwis...
<p>Comparison of kernelPLS with four other methods. For 5-fold cross validation classification accur...
<p>Comparison of kernelPLS with four other methods. For 10-fold cross validation classification accu...
<p>For each phenotype (column 1), the optimal mode of classification (“BAGS” or “C”) (second column)...
<p>Comparison of kernelPLS with four other feature filters for the classification accuracy(%) and AU...
(a-i) SVM, (b-i) Pin-SVM, (c-i) TWSVM, (d-i) THSVM, (e-i) QHSVM (τ = 0) and (f-i) QHSVM (τ = 0.5), w...
<p>Comparison of the performance of Support Vector Machine (SVM) classifier with sampling using poly...
<p><i>(A</i>,<i>B)</i> The <i>p</i>-values (color scale shown, adjusted by false discovery rate cont...
<p>The relative measure of model performance, i.e. the per-bin log-likelihood Δ<i>p</i> (see <a href...
<p>Comparison of classification results obtained through 5-fold cross validation with respect to dif...
Performance of sparse and non-sparse discriminant models in internal validation compared on the same...
<p>Each panel compares performance of different regions for a different combination of rank (genus o...
<p>The correct rates (%) were derived with systematically varying number of labels (L), number of sa...
<p>Each boxplot summarizes Kappa for classifying the metabolic inheritance patterns (mIPs) from 41 e...
<p>Comparison of prediction accuracy on four multiclass classification datasets by varying the numbe...
<p>For each subgraph, no significant difference in pairwise comparisons is marked with “~”, otherwis...
<p>Comparison of kernelPLS with four other methods. For 5-fold cross validation classification accur...
<p>Comparison of kernelPLS with four other methods. For 10-fold cross validation classification accu...
<p>For each phenotype (column 1), the optimal mode of classification (“BAGS” or “C”) (second column)...
<p>Comparison of kernelPLS with four other feature filters for the classification accuracy(%) and AU...
(a-i) SVM, (b-i) Pin-SVM, (c-i) TWSVM, (d-i) THSVM, (e-i) QHSVM (τ = 0) and (f-i) QHSVM (τ = 0.5), w...
<p>Comparison of the performance of Support Vector Machine (SVM) classifier with sampling using poly...
<p><i>(A</i>,<i>B)</i> The <i>p</i>-values (color scale shown, adjusted by false discovery rate cont...
<p>The relative measure of model performance, i.e. the per-bin log-likelihood Δ<i>p</i> (see <a href...
<p>Comparison of classification results obtained through 5-fold cross validation with respect to dif...
Performance of sparse and non-sparse discriminant models in internal validation compared on the same...
<p>Each panel compares performance of different regions for a different combination of rank (genus o...
<p>The correct rates (%) were derived with systematically varying number of labels (L), number of sa...
<p>Each boxplot summarizes Kappa for classifying the metabolic inheritance patterns (mIPs) from 41 e...
<p>Comparison of prediction accuracy on four multiclass classification datasets by varying the numbe...
<p>For each subgraph, no significant difference in pairwise comparisons is marked with “~”, otherwis...