fix ref

book/chapters/chapter6/feature_selection.qmd

@@ -405,7 +405,7 @@ The trained model can now be used to make a prediction on external data.
 
 ### Optimizing Multiple Performance Measures {#sec-multicrit-featsel}
 
-You might want to use multiple criteria to evaluate the performance of the feature subsets. With `mlr3fselect`, the result is the collection of all feature subsets which are not Pareto-dominated\index{Pareto optimality} by another subset. Again, we point out the similarity with HPO and refer to multi-objective hyperparameter optimization [see @sec-multi-metrics-tuning and @karl2022].
+You might want to use multiple criteria to evaluate the performance of the feature subsets. With `mlr3fselect`, the result is the collection of all feature subsets which are not Pareto-dominated\index{Pareto optimality} by another subset. Again, we point out the similarity with HPO and refer to multi-objective hyperparameter optimization (see @sec-multi-metrics-tuning and @karl2022).
 
 In the following example, we will perform feature selection on the sonar dataset. This time, we will use `r ref("FSelectInstanceMultiCrit")` to select a subset of features that has high sensitivity, i.e. TPR, and high specificity, i.e. TNR. The feature selection process with multiple criteria is similar to that with a single criterion, except that we select two measures to be optimized: