diff --git a/src/psycopt2d/config/preprocessing/default_preprocessing.yaml b/src/psycopt2d/config/preprocessing/default_preprocessing.yaml
index 3abc70b5..10aa9d6a 100644
--- a/src/psycopt2d/config/preprocessing/default_preprocessing.yaml
+++ b/src/psycopt2d/config/preprocessing/default_preprocessing.yaml
@@ -1,8 +1,8 @@
 convert_to_boolean: false
 convert_datetimes_to_ordinal: false
-imputation_method: "most_frequent" # (str): Options include 2most_frequent"
+imputation_method: "most_frequent" # (str): Options include "most_frequent"
 transform: null # (str|null): Transformation applied to all predictors after imputation. Options include "z-score-normalization"
-feature_selection:
-  name: null
+feature_selection: 
+  name: null # (str) Options include "f_classif", "chi2" and "mutual_info_classif". Default to use is "mutual_info_classif".
   params:
     percentile: 10 # (int): Percent of features to keep. Defaults to 10.
diff --git a/src/psycopt2d/train_model.py b/src/psycopt2d/train_model.py
index b98a4dc2..d095b07e 100644
--- a/src/psycopt2d/train_model.py
+++ b/src/psycopt2d/train_model.py
@@ -9,7 +9,12 @@
 import wandb
 from omegaconf import OmegaConf
 from omegaconf.dictconfig import DictConfig
-from sklearn.feature_selection import SelectPercentile, chi2, f_classif
+from sklearn.feature_selection import (
+    SelectPercentile,
+    chi2,
+    f_classif,
+    mutual_info_classif,
+)
 from sklearn.impute import SimpleImputer
 from sklearn.metrics import roc_auc_score
 from sklearn.model_selection import StratifiedGroupKFold
@@ -90,6 +95,16 @@ def create_preprocessing_pipeline(cfg: FullConfigSchema):
                 ),
             ),
         )
+    if cfg.preprocessing.feature_selection.name == "mutual_info_classif":
+        steps.append(
+            (
+                "feature_selection",
+                SelectPercentile(
+                    mutual_info_classif,
+                    percentile=cfg.preprocessing.feature_selection.params["percentile"],
+                ),
+            ),
+        )
 
     return Pipeline(steps)