A Multicriteria Approach to Find Predictive and Sparse Models with Stable Feature Selection for High-Dimensional Data

<div>Pareto front for the Pareto optimal configurations for data set <span class="monospace">AP_Colon_Kidney</span> (a) and <span class="monospace">Stomach</span> (b) considering the mean number of chosen features and the stability value. Only configurations whose mean misclassification rate is within 0.05 of the best classification accuracy are taken into account. Additionally, the accuracy optimal configurations are shown.</div>

Computational and Mathematical Methods in Medicine

fig10

Figure 10

Figure 10: A Multicriteria Approach to Find Predictive and Sparse Models with Stable Feature Selection for High-Dimensional Data 