Update app.py
Browse files
app.py
CHANGED
|
@@ -70,14 +70,14 @@ with gr.Blocks(title=title) as demo:
|
|
| 70 |
gr.Markdown(f"# {title}")
|
| 71 |
gr.Markdown(
|
| 72 |
"""
|
| 73 |
-
|
| 74 |
To showcase the difference between the two methods, we add two random features to the Titanic dataset. \
|
| 75 |
The first random feature is categorical and the second one is numerical. \
|
| 76 |
The categorical feature can have its number of categories changed \
|
| 77 |
and the numerical feature is sampled from a Standard Normal Distribution. \
|
| 78 |
Random Forest hyperparameters can also be changed to verify the impact of model complexity on the feature importances.
|
| 79 |
|
| 80 |
-
[
|
| 81 |
"""
|
| 82 |
)
|
| 83 |
|
|
|
|
| 70 |
gr.Markdown(f"# {title}")
|
| 71 |
gr.Markdown(
|
| 72 |
"""
|
| 73 |
+
This demo compares the feature importances of a Random Forest classifier using the Mean Decrease Impurity (MDI) method and the Permutation Importance method. \
|
| 74 |
To showcase the difference between the two methods, we add two random features to the Titanic dataset. \
|
| 75 |
The first random feature is categorical and the second one is numerical. \
|
| 76 |
The categorical feature can have its number of categories changed \
|
| 77 |
and the numerical feature is sampled from a Standard Normal Distribution. \
|
| 78 |
Random Forest hyperparameters can also be changed to verify the impact of model complexity on the feature importances.
|
| 79 |
|
| 80 |
+
See the original scikit-learn example [here](https://scikit-learn.org/stable/auto_examples/inspection/plot_permutation_importance.html#sphx-glr-auto-examples-inspection-plot-permutation-importance-py).
|
| 81 |
"""
|
| 82 |
)
|
| 83 |
|